a
HUMAN VISUAL-VESTIBULAR INTERACTIONS
DURING POSTURAL RESPONSES TO BRIEF FALLS
by
ROGER WILLIAM WICKE
S.B.
in Electrical
Encineering,
Massachusetts Institute of Technology
(1975)
S.M. in Electrical Engineering,
Massachusetts Institute of Technology (1976)
SUBMITTED IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
at the
MASSACHUSETTS
INSTITUTE OF TECHNOLOGY
June 1980
Massachusetts
Institute of Technology 1980
Signature of Authc
&epartment of Electbical
Engineering
June 1980
Certified by
Thesis 6'dperfyidckr
Accepted by
Chairman, Committee on Biomedical Engineering
2-
HUMAN VISUAL-VESTIBULAR INTERACTIONS
DURING POSTURAL RESPONSES TO BRIEF FALLS
by
ROGER WILLIAM WICKE
Submitted to the Committee on Biomedical Engineering
on 9 June 1980 in Partial Fulfillment of the
Requirements for the degree of Doctor of Philosophy
in the field of Biomedical Engineering
ABSTRACT
Human subjects were suspended in a safety harness 28 cm
above the floor by a steel cable connected to a computer
After
(electromagnetic brake).
controlled force generator
were unexpectedly, released, various controlled
the subjects
patterns of downward acceleration (less than one g) could be
EIG activity was recorded
During the falls,
produced.
tibialis
simultaneously from the gastrocnemius, soleus,
knee
with
anterior, rectus femoris, and biceps femoris, along
and ankle joint angle in one leg. Subjects were tested eyes
both in darkness and in light
closed and also eyes open,
using a wide field visual display. The display scene could
be moved downwards' at exactly the same velocity as the moving
("normal
subject, left fixed with respect to the laboratory
or moved upwards at a speed equal to the
field"),
visual
Ten
subject's falling speed ("upward moving visual field").
45
of
a
total
underwent
each
vestibularly normal subjects
drops, experiencing three replications of each vision/motion
combination used.
both short and
Under normal visual field conditions,
long ' latency postural responses were seen, which were
dependent on the magnitude of the acceleration stimulus.
of the visual conditions significantly altered both
Certain
the short and long latency responses in most of the muscles
in the
prominent
were
particularly
Effects
tested.
pronounced
and were also more
gastrocnemius and soleus,
The upward moving visual field
(0.5G) falls.
during slow
in
the short latency EMG reaction
condition increased
gastrocnemius and soleus for 0.5G falls. A preliminary model
EMG
for visual-vestibular interaction in short latency
Long latency responses are more
responses is presented.
variable and are not conducive to a simple interpretation.
Thesis Supervisor: Charles M. Oman
Title: Helmholtz Associate Professor of Aeronautics
Principal Research Scientist
and Astronautics;
-3-
MEMBERS OF DOCTORAL ADVISORY COMMITTEE
Charles M. Oman, Ph.D
Helmholtz Associate Professor; Principal Research Scientist
Department of Aeronautics and Astronautics
Massachusetts Institute of Technology
Laurence R. Young, Sc.D.
Professor
Department. of Aeronautics and Astronautics
Massachusetts Institute of Technology
Sheldon R. Simon, M.D.
Associate in Orthopedic Surgery,
Director of Gait Laboratory, Children's Hospital,
Boston, Massachusetts;
Lecturer, Department of Mechanical Engineering
Massachusetts Institute of Technology
Douglas G. D. Watt, M.D.,
Assistant Professor
Department of Physiology
McGill University
Ph.D.
-4-
ACKNOWLEDGEMENTS
for
Oman and Young
I would like to thank Professors
evolved
which
complete a project
to
freedom
the
providing
laboratory
previous
the mainstream of
somewhat outside
After my obsession with vestibulospinal reflexes
projects.
valuable
they contributed
became evident,
falls
during
suggestions and steered me away from unworkable methods.
examining human
in
time
contributed valuable
Simon
Dr.
proper safety
of
determination
the
in
aided
and
subjects
precautions.
the
in discussing both
very helpful
Professor Watt was
technique and the physiology
EMIG recording
of
limitations
underlying the experiment.
clinical
performed
Weiss
Alfred
Dr.
and
Tole
John
Dr.
vestibular examinations of human subjects who participated in
the experiments.
Luca
Carlo de
Discussions with Dr. Neville Hogan and Dr.
practical
of
were
processing
and
recording
EMG
concerning
the
and designing
importance in developing my technique
necessary equipment.
reviewed my proposal
Dr. Joseph Bauer and Dr. Alan \Natapoff
greatly
design, which
statistical
experimental
the
for
increased the value of the final data.
I am also indebted to the
endured the experiments.
many human subjects
who patiently
Roger William Wicke
This project was supported by the National Aeronautics and
under grant
Center,
Space Administration, Ames Research
NSG-2032.
number
supported by a
The author was
Fellowship, 1978-1980.
Health Science and Technology
-
5
TABLE OF CONTENTS
TITLE
PAGE.....................
AB STRACT.
. .
...
. .
1
. *...............*
..
ACKIIOWLEDGEMENTS.............
.........
............
2
.e.g.....
*** ***
*4
**
g...............*O
0
0
0
a
0
4
0
5
TABLE OF CONTENTS...............
CHAPTER 1
-
INTRODUCTION: UNEXPECTED FALLS AS A TOOL
FOR STUDYING VESTIBULOSPINAL MECHANISMS.
for choosing unexpected
16
1.1
Rationale
1.2
Brief description of procedure.............
19
1.3
Action of major leg muscles................
20
1.4
Salient features of the electromyographic
response..
. 00.
. 0
.
0
. ..
0
.
.
.
falls....
15
.
......
25
1.5
Effects of habituation and expectation.....
26
1.6
Preprogrammed aspects versus continually
updated parameters of the reaction......
28
1.7
Previous studies of unexpected falls.......
29
1.8
Possible visual-vestibular interactions....
32
CHAPTER 2
-DESCRIPTION
OF APPARATJS...................
40
2.1
Electric brake and controller..............
41
2.2
Harness and suspension system..............
44
2.3
Additional safety measures.................
46
2.4
Visual field motion.......................
47
2.5
EMG electrodes and preamplification........
52
2.6
EMG processing....................
53
2.7
Acceleration measurements..................
.....
55
6
-
2.8
Goniometers.................
2.9
Computer control of stimuli and data
conversion................ ... ..........
55
2.10 Miscellaneous equipment.....
CHAPTER 3
-
..........
57
00*
*0
*
0
59
..
62
EXPERIMENTAL PROCEDURE AND STATISTICAL
DESIGN ......
.
.
.
.
.
.
.
.
.
.
..
3.1
Preparation of the subject.... .............
62
3.2
Basic procedure for each drop..........
64
3.3
Statistical design of experiments..........
67
EXPERIMENTS
CHAPTER 4
AND RESULTS...............
71
4.1
Stimulus combinations....................
71
4.2
Preliminary tests ...................
75
4.3
Testing order and coding of trials .........
81
4.4
General description of the EMG response....
82
4.5
Statistical analysis of the data...........
4.6
Reactions to different
104
acceleration
131
4. 7
General habituation.
4. 8
Main visual effects.............
4. 9
Interaction
..............
00...0
142
effects............ .000.00.....
INTERPRETATIONS AND CONCLUSIONS......
5. 1
153
Proposed model for early response
generation...........
5.2
147
149
4. 1C Coactivation coefficients..................
CHAPTER 5
141
.
........
162
Possible physiological explanations for
visual-vestibular interactions during
the early response...................... 169
-7
5.3
Chapter 6
Additional experiments suggested
by the early response model.............
174
5.4
Synthesis of the late response............. 176
5.5
Significance of results in terms of
models of internal feedback during
intended movement.........
-
APPENDIX A -
SUMMARY..............................
179
184
.....
REVIEW OF POSTURAL TESTS OF
190
VESTIBULAR'FUNCTION.....................
A.l
Advantages and disadvantages of
behavioral tests ....................
.
190
Common stimuli used to provide gross
disturbances to the vestibular systen...
191
Individual limb movements induced by
vestibular stimuli......................
192
A.4
Normal and induced sway during standing....
194
A.5
The
A.6
Effects of vestibular disturbances on
A.2
A.3
A.7
APPENDIX B -
tipping test.........................
segmental reflexes....................
196
Tests of balance during walking............
197
THE FUNCTION OF THE VESTIBULAR SYSTEM
AND RELATED BRAINSTEM STRUCTURES
IN THE CONTROL OF POSTURE........
B.2
General structure and innervation of
...
the vestibular nuclei............ ...
.......
Afferent inflow to the vestibular nu
B.3
Primary vestibular afferents........
B.1
195
200
200
207
207
-8
B.4
Response patterns of central vestibular
neurons to primary vestibular
.
209
B.5
Commissural connections....................
214
B.6
Direct spinal afferents.....................
215
B.7
Cerebellar function in relation
to the vestibular nuclei..............
217
afferent input .............
..........
217
B.7.1
General cerebellar physiology............
B.7.2
Input to the vestibulo-cerehellum........ 218
B.7.3
Output of the vestibuld-cerebellum.......
221
B.7.4
Direct projections of the spinal
cerebellum to the vestibular nuclei.....
222
Fastiaial
B.7.5
B.8
nuclei..
223
Reticular formation afferents to the
225
vestibular nuclei............... .......
B.9
Efferents
from vestibular nuclei to
reticular formation.....................
226
B.10 Influence of brainstem and cerebellum
on spinal mechanisms................*...
226
B.11 Lateral vestibulospinal tract (lVST)....... 227
B.12 Medial vestibulospinal
tract
(mVST)........ 229
B.13 Caudal vestibulospinal tract...............
230
B.14 Reticulospinal tract.....................
230
B.15 Control of the
motor system.............
231
B.16 Functional characteristics of stretch
refl ....
x... ... ...
B.17
o
....
...
2 interaction during reflexes
and movement..... . . . .. ... . ... ....
....
...
236
...
238
-
. . ..
B.18 Effects of LVT1 and VST activity on
motor activity initiated elsewhere...... 242
B.19 Flexor reflex interaction with the
vestibular system.......
...............
244
-9
B.20 Sources of decerebrate rigidity............
246
B.21 Eighth nerve section in intact animals..... 247
B.22 Fastigial lesions and atonia............... 249
B.23 Brainstem and cerebellar function during
locomotion ............
.
.
.
.
.
.
..
.
250
APPENDIX C -
EMG PROCESSOR SCHEMATIC AND
SPECIFICATIONS..................0 254
APPENDIX D -
STATISTICAL ANALYSIS
PROCEDURE .
.............
.
258
263
l 1u
LIST OF FIGURES AND TABLES
Fig.
1.1
- Major muscles of the hip, thigh and leg...
Fig.
1.2
-
Fig.
2.1
-
Idealized limb positioning at the instant
of contact with the ground
following a short fall..................
43
2. 2
-
Braking system............
Fig.
2. 3
-
Visual field motion svstem...........
Fig.
2. 4
- Example of type of visual field
...............
pattern used. ...................
-
42
.
Fia.
2.5
21
Subject suspension system and safety
devices..0 ........... ...............
Fig.
21
45
.....
49
......
Map of restricted visual field using
0 0 0 *
. . .. 0 . 0 . * .. 04 .......
goggles,.... ....
51
Fig.
2. 6
- Electrode preamplifier schematic..........
54
Fig.
2. 7
- Accelerometer and amplifier schematic.....
56
Fig.
2. 8
- Leg goniometer assembly...................
58
Fig.
3.1
- Leg electrode positions...................
65
Table 4.1
- Description of subjects...................
72
Fig.
-
4.1
Acceleration and velocity profiles
for acceleration stimuli used
in drop tests .......................
Table 4.2
-
.
Combinations of acceleration profile
and visual field which were tested......
Fig.
4.2
Table 4.3
- Acceleration measurements taken at the
harness with 160 lbs. of added weight
for calibration........ ................
-
73
76
78
Drop test sequence for block 1
and block 2 (reverse of block l)........
83
-
Table 4.4
ILJI
Drop tests grouped by test type,
giving sequence positions
....
in blocks 1 and 2..........
...
0.
84
Fig.
4.3.1
-
Subject EMA, tests 0.5G/N/Bl/1-3 .........
85
Fig.
4.3.2
-
Subject
EMA,
tests 0.85G/N/Bl/1-3.........
86
Fig.
4.3.3
-
Subject
EMA,
tests RG/N/Bl/l-3...000000000
87
Fig.
4.3.4
-
Subject
EMA,
tests
OG/D/Bl/-3.0000000000
88
Fig.
4.3.5
-
Subject
DDL,
tests 0.5G/N/Bl/l-3.......0.
89
Fig.
4.3.6
-
Subject
DDL,
tests 0.85G/N/Bl/l-3 .....
90
Fig.
4.3.7
-
Subject
LLA,
tests 0.5G/N/B2/l-3..........
91
Fig.
4.3.8
-
Subject
LLA,
tests 0.85G/N/B2/-3.........
92
Fig.
4.3.9
-
Subject
TWA,
tests 0.5G/N/B2/l-3 ..........
93
Fig.
4.3.10 - Subject
TWA, tests 0.85G//B2/l-3.........
94
Fig.
4.3.11 -
Subject
PPE,
tests 0.5G/N/B2/1-3 ..........
95
Fig.
4.3.12 -
Subject
PPE,
tests 0.85G/N/B2/l-3. .....
.
96
Fig.
4.3.13 -
Subject
tests 0.5G/N/Bl/l-3 ..........
97
Fig.
4.3.14 -
Subject
Fig. 4.4.1
Fig.
4.4.2
-
PIlE,
tests 0.85G/N/Bl/l-3.
Gastrocnemius,
integrated EMG during entire
0...0
...00000
fall....
000
98
108
Soleus,
integrated ENG during entire fall....... 109
Fig.
4.4.3
-
Tibialis anterior,
integrated EMG during entire fall....... 110
Fig.
4.4.4
-
Rectus femoris,
integrated EMG during entire fall....... 111
Fig.
4.4.5
-
Biceps femoris,
integrated EMG during entire fall.......
Fig.
4.4.6
-
112
Gastrocnemius,
integrated EMG during early response.... 113
Fig.
4.4.7
-
Soleus,
integrated EMG during early response.... 114
- 12
Fig.
4.4.8
Fig.
4.4.9
-
- Tibialis anterior,
integrated EMG during early response....
-
Rectus
115
femoris,
integrated EMG during early response.... 116
Fig.
4.4.10 - Biceps femoris,
integrated EMG during early response .... 117
Fig. 4.4.11 -
Gastrocnemius,
integrated EMG during late response..... 118
Fig.
4.4.12 -
Soleus,
integrated EMG during late response.....
Fig. 4.4.13 - Tibialis anterior,
integrated EMG during late response.....
Fig. 4.4.14 -
Fig.
Fig.
4.4.16 4.4.17 -
120
Rectus femoris,
integrated EMG during late response.....
Fig. 4.4.15 -
119
121
Biceps femoris,
integrated EMG during late response.....
122
Gastrocnemius,
integrated EMG( during contact...........
123
Soleus,
integrated EMG during contact........... 124
Fig. 4.4.18 -
Tihialis anterior,
integrated EMG during contact... .. a.s..
Fig. 4.4.19 -
.
125
Rectus femoris,
integrated EMG during contact...........
126
Fig. 4.4.20 - Biceps femoris,
integrated EMG during contact...........
127
Fig.
4.5.1
-
Integrated EMG during early response,
normal visual field conditions.......... 132
Fig.
Fig.
4.5.2
4.5.3
-
Integrated EM4G during late
response,
normal visual field conditions.......06
Average integrated EPMG
136
(per unit time)
during early response,
normal visual field conditions..........
Fig.
4.5.4
-
137
Integrated EMG during contact,
normal visual field conditions.......... 139
- 13
Table 4.5
Table 4.6
Fig.
5.1
-
-
-
Coactivation coefficients for 0.5G/N
and 0.85G/N falls.................
150
GN-TA and SL-TA coactivation coefficients
for 0.5G and 0.85G late responses:
all visual conditions.................
150
Model for determination of early response
magnitude..................... ........
. 163
201
Table B.1
-
Abbreviations used in text................
Fig. B.1
-
Primary vestibular afferents
to cerebellum..... ...........
.
..
Secondary vestibular afferents
to cerebellum....
..............
.
...
Fig.
Fig.
B.2
B.3
-
-
*.
Cerebellar outflow to the vestibular
nuclei ...... . . .
..
. . . .
.
projections....
....
.....
.
.
B.4
-
Fig.
B.5
- Simplified representation of VST and MLF
innervation pattern.................
Fig.
C..
Table C.1
Fig.
Fig.
C. 2
C.3
tract
202
.
203
..
. a 04
Fig.
Vestibulospinal
.
205
205
- EMG processor schematic for one channel... 255
-
Transfer functions for individual stages
of the EMG processor....................
256
- Magnitude plot for bandpass stage of EMG
processor ...
..............................
257
-
Step response of third order averager
compared to Butterworth filters.........
Table D.l
204
-
257
Variance estimates for main and
interaction terms of three way split
plot experimental design................. 261
-
14
-
-
I
4M
:)
CHAPTER 1
INTRODUCTION:
UNEXPECTED FALLS
AS A TOOL FOR STUDYING
VESTIBULOSPINAL MECHANISMS
ear,
The labyrinth organs of the inner
the
and otoliths, provide acceleration
canals
semicircular
extremely
and gravitational orientation information which is
for
develop
labyrinthine
balance,
maintaining
in
difficulty
cases
in
However,
proportions.
suffer
initially
disorders
People
stability.
postural
maintaining
useful
include
which
who
from
which may reach severe
permanent
of
vestibular
patients are often capable of compensating
for the
loss of vestibular function by increasing the role of
vision
disorders,
and proprioception in postural maintenance.
This phenomenon raises the question of what role vision
might
have
people.
in
Recent
conflicting
contributing to postural reactions in normal
reports
visual
cues
have
may
indicated
in
that
induce destabilizing
postural
reactions with very short latencies (Lestienne et al.,
Nashner
baboon).
and
Berthoz,
Therefore,
1978;
Vidal
et
al.,
one problem is to determine
man,
1977;
1979,
in the
the
manner
in which visual cues are combined with vestibular information
under various circumstances.
to
-
Because the motor system is designed
far
stable posture, motor commands arising
must be coordinated
system
accomplish
of tasks than merely the maintenance
variety
greater
to
to the current state of activity.
A wide variety of motor activities seem to
be
organized
influence
vestibular
these
unit recording and of electrical
reveal
tracts
motor
muscle
similarly organized.
be
to
expected
upon
groups
of
stimulation
in locomotion (Grillner et al.,
1975; Kots, 1976; Shik
posture
.of
and
vestibulospinal
is
Orlovsky,
of
goal
visual-vestibular
this
one
mechanisms,
interactions
two
fallI
is
study
during
Sensory interaction and
are
if
example,
vestibulospinal
of
Rational?. for choosinZ unexpected
adjustments.
For
pages 227-252.
see Appendix B,
synthesis
initial
The
the distribution of commands is asymmetric.
flexed,
One
1976).
the subject may also influence which muscles are
For a more extensive review
1.1
limb,
the
1970, 1971; Hongo et al.,
affected by vestibulospinal commands.
leg
be
descending
activity with respect to extension and flexion of
as
might
The results of single
coordinated
highly
in
exertion of
of synergistic muscle activation patterns;
terms
of
vestibular
the
from
a
to
rapid
investigate
postural
vestibulospinal
motor
aspects of the problem which are treated
simultaneously in this study.
17
-
With this goal,
the
first
major
Investigations of
reactions.
of
disturbances
external
reactions
postural
human
to
position dur ing stance have
body
been performed by Nashner (1970,
1971,
1976,
1973,
1977).
motions were later added to the basic paradigm
field
(Nashner and Berthoz, 1978),
which
of
consisted
asking
a
maintain upright stance on a platf orm which could
to
person
of
an appropriate paradigm for studying rapid postural
choosing
Visual
consisted
task
feedback
angle
ankle
be servo-controlled to eliminate
and
could also be suddenly moved in an anteroposte rior direction.
While this type of experiment is conducive to
analysis
that
otoliths,
canals,
semicircular
makes
information
interpretation
In
difficult.
vision
and
difficult to extract any
test makes it
all
contribute
physiological
fundamental
addition,
hip,
and
ankle
the
at
sensors
proprioceptive
system
the fact
a response to an external disturbance,
of
significant
linear
the nature of this
kind
of
information
about sensory to motor processing and transmission latencies.
were
Unexpected, short falls
external
in
disturbance
advantageous features.
defined,
allowing
the
fall
after landing.
chosen
as
the
study because of a number of
The onset of the stimulus is
various
sensory-motor
latencies
clearly
to
be
The primary effect of postural reactions
clearly determined.
during
this
finally
is
to
minimize jolts and loss of balance
Unlike the case of disturbance during
stance
in which the reaction itself immediately affects the stimulus
18 -
-
to the vestibular organs in a closed-loop feedback mode,
the
postural reactions during a fall are not part of a corrective
feedback scheme.
employed.
Feedback
overlooked,
release
Instead,
though;
may
alter
of
a
a
muscle
strategy
nature
must
be
not
be
must
contractions
following
head position, and, thus, alter the
effective stimulus to the
gamma
different
neck
the
predictive
labyrinths.
Activation
of
neck
motoneurones-may result in even faster changes in neck
afferent feedback reaching vestibulosninal tract neurons.
addition,
since' the
mass
of
the
legs
is
In
a significant
fraction of body mass, knee and hip flexion will also
result
in motion of the upper body and head which would be different
from that of a passive body.
By surrounding the subject with a movable visual field,
the
visual
cues
can
be
altered
independently
of
the
accelerations
and
acceleration stimuli.
By confining the study to
vertical
visual
field
vertical
motions, any significant effects can
probably be explained in terms of gravireceptor
without
the
influence.
of
the
confounding
effects
of
involvement,
semicircular
(If the head is facing forward during
fall,
antero-posterior
head
initiation
motion due to passive
effects was observed to be minimal.) Experimental
of
the
linear
and anqular modes is important,
influences from the canals
and
canal
otoliths
may
separation
since spinal
be
channeled
19 -
-
pathways,
different
through
and intermediate processing of
angular and linear accelerations may be totally different
(Precht,
character
1974).
The remainder of this chapter consists of a
of
underlying
rationale
the
the
discussion
development of the basic
The action
procedure, with references to the literature.
of
major leg muscle groups is reviewed in the context of an
the
appropriate reaction to a short fall, followed by
of
in
reactions
the
After the essential features of
are
falls
summary
EMG reactions during falls as reported by other
observed
authors.
a
to
outlined, individual articles from the literature
detail.
are examined in more
the
Throughout
review,
the
findings are discussed with emphasis on their significance in
motivating certain aspects of the present study.
1.2
Brief description of procedure
Each
subject
was
suspended
in
a
safety
harness
connected by a cable to an electric brake and surrounded by a
patterned visual field which could be moved vertically.
subject
dropped unexpectedly at accelerations of 0.85g
was
(somewhat slower than free fall)
field
could
be
or
less,
and
varying
to
determine
adjusting
the
visual
moved either up, down, or remain stationary
with respect to the laboratory during the fall.
by
The
In addition,
the acceleration in mid-fall, it may be possible
the
if
the
motor
sensorimotor
response
system
timing
to
is
capable
account
of
for the
-
6- W
-
alteration.
1..3.
Action .of maigr 1.g muscles
In this study, EMG measurements from
the
were
leg
soleus,
biceps
femoris.
femoris
and
anterior,
tibialis
In
order
of
these
muscles
will
to
action
interpretation of the EMG recordings, the primary
each
of
These
recorded using surface EMG electrodes.
muscles were the gastrocnemius,
rectus
muscles
five
aid
of
be reviewed, as well as other.
important muscles of the hip and leg which are not accessible
with surface recording techniques.
Figure 1.1 on page 21 shows a
the
groups of the leg.
muscle
major
version
schematized
of
The ideal strategy of
muscle activation to prepare for landing involves both proper
positioning
of
the
stiffness about each
minimize
joints
joint
and adjustment of the effective
during
order
In
landing.
to
transmission of impact forces of the landing to the
upper part of the body and spinal column,
should be flexed,
a position,
the
hip
and
as shown in figure 1.2 on page 21.
knee
In such
the weight and momentum of the body during impact
will tend to cause further passive flexion.
EMG recordings do not
allow
conclusions
quantitative
about the associated muscle forces; only qualitative features
are indicated in comparisons of the recordings
muscles.
of
different
With this in mind, some of the constraints involved
-
21
gluteal muscles
iliopsoas
hamstrings,
including
biceps femoris
quadriceps:
rectus femoris
vastus medialis
vastus intermedius
vastus lateralis
\
short head of
biceps femoris
gastrocnemius
-
soleus
'< tibialis
anterior
Figure 1.1.
Major muscles of the hip, thigh and leg.
hip
knee
ankle
Figure
Idealized limb positioning at the instant of
1.2.
contact with the ground following a short fall.
-
22
-
in possible landing strategies should be considered.
It is obvious that if the hip is
the knee is fully extended
(A =0),
( "=0)
extended
and
then the impact of landing
on the upper body can only be reduced by plantar flexing
feet
such
0
that
>0 at contact.
Contraction of the soleus
muscle at contact will slow the passive dorsiflexion
ankle
in
order
to
with
if
of
the
reduce the velocity of the heel when it
contacts the floor a fraction of a second later.
fall,
the
During
the
the tibialis anterior is contracted simultaneously
the
the
soleus,
ankle
position
may
change
not
significantly, but the effective stiffness will increase.
the knee and hip are extended, the forces of contact
heel
will
be
transmitted
bones and spinal column,
directly
If
at
the
upward through the leg
analogous to a rigid pillar
hitting
the ground.
For
strategy
even
short
would
falls,
the
previous
be painful if not dangerous.
hypothetical
By flexing the
knee prior to contact, much of the imp-act can be absorbed
further
passive
flexion.
In
addition, all of the muscles
acting at the knee joint can then exert fine control
landing
with
varying
degrees
of
contraction.
gastrocnemius muscle acts at both the knee and ankle
contraction
of
the
Since the
joints,
will achieve the same effect as the soleus as in
the prior strategy, but will also act to flex the knee.
hamstring
by
muscles,
The
including the biceps femoris also aid in
- 23 -
knee flexion.
Hip flexion prior to contact will
peak
forces
acting
on
the
muscle is the most effective
rectus
spinal
of
reduce
of
to
the
hip,
the
the
The iliopsoas
femoris also contributes to hip flexion.
but
the
Because the
intra-abdominal
the iliac pelvis and spine, it is inaccessible with
surface EMG techniques.
the
column.
flexor
iliopsoas is a deep muscle attached
face
further
The rectus femoris lies just beneath
fascia lata of the anterior thigh and is easily recorded
with surface electrodes.
contributes
to
knee
However, the
extension
rectus
since
femoris
also
it
is
part
of the
Since the biceps femoris, in addition
to
flexing
quadriceps group.
knee,
and
also
extends
the hip, recordings from rectus femoris
biceps
femoris
do
concerning
hip
not
flexion.
biceps
allow
Many
flexion and extension, so
more
the
that
accurate
other
any
conclusions
muscles
results
affect
obtained
hip
from
femoris and rectus femoris must be interpreted from a
phenomenological
soleus,
and
tibialis
ankle joint, so that
perspective.
anterior
more
The
gastrocnemius,
are the prime movers of the
definite
indications
concerning
mechanical strategies should be evident from EMG recordings.
The previous discussion has
with
proper
limb
positioning
been
in
flexion of the hip, knee, and ankle
concerned
order
joints
to
primarily
allow passive
during
contact.
24 -
-
It
be pointed out that during suspension in a safety
should
slightly
harness, the knee and hip are usually
flexed
when
relaxed; any additional flexion may be achieved by relatively
limbs
unloaded.
are
By
the
contrast,
forces
prevent collapse following contact are much
the
the
because
muscles
correct
low level contractions of the
needed to
greater
due
to
sudden loading of the legs by the weight and momentum of
the body.
1_4
Salient features of the electromyographic
Electromyographic (EMG)
manifestations
lower
leg
patterns
patterns
response
involuntary
an
of
the
stereotyped
EMG
These
falls.
unexpected
during
demonstrate
eye,
neck and around the
in
other muscles including those of the
numerous
and
muscles,
reactions
response
of
include
vestibular
origin, and as such, they provide an excellent medium for the
The
stimuli.
motor
and
function
study of otolith
in
reaction
involuntary
begins about 75 msec after release and
Greenwood
100 msec.
and
responses
phasic
soleus muscle
the
continues
(1976a)
Hopkins
to
for
about
found that this
reaction was absent in the two labrynthine defective patients
which
they
baboons,
Lacour
tested.
this
subcomponents,
et
al.
be
(1978) showed that in
separated
two
could
the first
of which disappears after bilateral
labyrinthectomy and the second of which is
magnitude.
into
response
These
only
reduced
in
results indicate that the labyrinths, and
- 25
-
specifically,
presumably
the
otoliths
responsible
for
triggering
the
early
primarily
are
following
response
release, though other sensory systems may contribute.
the
If the duration of the fall is long enough,
first
response is followed by a period of diminished EMG, and then,
a subsequent increase which reaches a maximum 40 to 140
prior
reaction is presumed
This latter
to contact.
upon
voluntary origin and is dependent
judgement
of
If
acceleration.
of
to be of
subject's
height above the landing platform and his
his
downward rate
the
both
msec
the
subject
releases
himself, no initial involuntary or "startle" reaction is seen
(Greenwood and Hopkins, 1976a).
responses
preprogrammed
The EMG reactions seem to be
similar
to
the
EMG burst
second
during unexpected falls.
1.5
Effect§ _f habituation and expectation
The
EMG
initial
(Greenwood
and
1976a).
Hopkins,
very
rapidly
Amplitudes
decrease
habituate
responses
significantly over the first several drops, but persist at
relatively
steady
level
thereafter,
a
for drops of the same
height and acceleration.
In this study the ordering of different test conditions
was
Therefore,
randomized.
habituation
significant.
to
the
For
drops,
example,
the
in
addition
testing
order
to
overall
may
be
the results of a given test may
-
depend
problems
on
the
were
aware of such
parameters
no-t
26
-
of
the
insuperable,
potential
but it
interactions
overall experimental design.
previous
test.
These
was necessary to be
for
(See Chapter 3.)
the
purpose
of
27
-
1.6
Preprogrammed
aspects
continually
versus
updated
parameters of the reaction
Many
a
It
is
of central concern to determine what
features of the motor sequence can
For
(1976b),
timing
according
example,
be
to
modified
during
Hopkins
and
Greenwood
the
seems that for falls of sufficient duration,
it
the
of
second,
EMG
functional
the
and
ground
his
rate
of
his
height
acceleration.
It was
hypothesized that the magnitude of the initial
during
acceleration
release
in
transient
is the cue that determines the
timing of the EMG prior to landing.
to
the
prior to
response
landing depends upon the subject's perception of
above
be
of
preprogrammed.
fall.
to
appear
jump
voluntary
aspects
One
to
way
test
this
the pattern of acceleration during the
would
be
change
fall.
Instead of an initial step in acceleration,
either
a
gradual increase or a series of smaller steps in acceleration
could be achieved with an appropriate sequence of restraining
on
forces
the
method, falls of
acceleration
If
harness suspension during the fall.
By this
radically
different
equal
duration,
but
sequences could be compared.
acceleration
changes after release can be
affect the response, then it
shown
is clear that there must exist a
time after which acceleration information from
the
otoliths
can no longer modify the timing of the pre-contact EMG.
latency might be comparable to the
involuntary
response,
but
to
not
latency
of
necessarily,
the
This
initial
since the two
28
-
-
transmission
neural
latencies reflect processing and
times
for different phenomena.
The
fact
unexpected
that
the
response
earliest
commands
vestibulospinal
indicate
to
be
to the soleus
transmitted
that galvanic stimulation of the vestibular
onset
the
of
as
H-reflex
the
apparatus causes potentiation of
following
for
time
minimum
Soleus H-reflex tests in humans by Kots
muscle, for example.
30 msec
an
release occurs at a latency of 75 msec should not
be used to infer that this represents the
(1976)
following
soon
as
Studies by
the stimulus.
Matthews and Whiteside (1960) in the human and by Watt in the
cat
personal
(1980,
communication)
revealed effects on the
but the
H-reflex with a similarly short latency,
initially
followed by a period of facilitation.
inhibitory,
It was hypothesized that
manifestation
of
was
effect
the
EMG
earliest
reaction
is
a
this facilitatory period, since the timing
of the two events coincides.
These latter results cannot
be
reconciled with those of Kots, but it is clear that the early
EMG reaction does not reflect the earliest activity
occuring
at the segmental level.
1.7
Previous studies of unexpect-gd falls
Melvill Jones and Watt (1971a) tested unexpected
by
electromagnetically
which a subject held.
releasing
an
They demonstrated
overhead
that
falls
handle onto
the
reaction
during landing could not be explained as a functional stretch
-
29
-
late
reflex, since such a reflex would have occurred too
be
(See Appendix B,
assistance- in preventing collapse.
of
to
section B.16, for a review of the functional stretch reflex.)
By dropping subjects from different heights, they showed that
contact;
were
for
because
jolt
uncomfortable
an
by
for a functional response to occur.
time
the
latency
landing
occuring
independently
of
a
with
74 msec
of
latency
If a
The short
should be independent of height.
response,
of
quality
functional stretch reflex were responsible, then the
of
to
falls of less than 160 msec duration, landings
characterized
insufficient
prior
determined
was
the gastrocnemius reaction timing
height, was also first documented by
drop
Melvill Jones and Watt.
Recent
of
experiments
personal
(1980,
Watt
communication) have been performed in the horizontal position
and in zero-g conditions in parabolic. aircraft
experimental
experiment,
subject
to
protocol
similar
was
The
to that in the previous
except that elastic bungee cords
the
flight.
connecting
the
landing platform replaced the gravitational
force; with the bungee cords under tension, the subject would
be
propelled toward the landing platform upon release of the
overhead handle.
the
early
The results from these tests indicated that
response
magnitude
horizontal position, over
a
is greatly reduced.
period
of
several
For the
hours
the
response gradually increases toward the magnitude seen in the
vertical, one-g situation.
This
suggests
that
the
one-g
- 30
of
biasing
the
early response
which
a potentiating effect on the
has
saccules
is
temporarily
diminished
when
the
saccules are unbiased.
In addition to substantiating the
Jones
and
(1976a,b)
defective
showed
that
magnitude is an
increasing
Their
subjects
were
cable
to
an
of
Melvill
and demonstrating the absence of the early
Watt,
response in labrynthine
Hopkins
results
the
of
and
Greenwood
early response
the
acceleration.
in a harness connected by a
suspended
By
soleus
function
electromagnetic
counterweights.
patients,
release
varying
the
and
system
a
of
counterweighting,
step
acceleration profiles of less than one-g could be tested.
variations
Several other
interesting
features
of
procedure
the
of the response.
If the subjects were
early
allowed to release themselve.s, no
revealed
response
occurred;
remained.
In
another variation, unexpected falls were tested in which
the
only
the
functional
eyes were closed and the fall height was chosen at
subject's
random from one of two heights;
with
response
pre-contact
the
late
reaction occurred
a latency which was between those for falls from the two
heights when the subject knew the height
tests
usually
elicited
sensations
of
beforehand.
surprise
from
These
the
subjects, indicating that the actual fall duration was either
greater
or
less
than expected.
This further verified that
the late EMG reaction is under voluntary
control
and
shows
- 31
none
of
the
features of a functional stretch reflex,
since
the timing can vary relative to the contact time in ambiguous
situations.
j.8
Possible visual-vestibular inter-actions
may
The initial involuntary reaction
useful
tool
for
studying vestibulospinal
other motor systems,
be generated
prove
provided that it
Its absence
people strongly
the
origin.
initial
reaction
semicircular
is.
as
its
still
canals,
present
in
1976).
Several
though.
For example, while peripheral
organs
vestibular
neurons,
interruption
vestibular nuclei,
vestibular
nuclei
possibilities
this
degeneration
may
and
vestibular
vestibulospinal
especially
motor
likely,
plugged
degeneration
(Watt,
considered,
of
of
primary
also
may
the
cause
and oculomotor pathways through the
may
act
input to trigger descending
visual
be
degeneration
cause
which
with
responsible
must
may
visual
of
The fact that this
cats
indicates that the otoliths are probably
vestibular
implicates
disappears after labyrinthectomy,
but
other
a
can be clearly shown to
in totally labyrinthine-defective
system
be
interactions with
primarily by the vestibular system.
vestibular
to
affect
as
motor
information
but
into
This
commands.
of visual-vestibular interaction
The
gates, requiring vestibular
activity,
considering
function.
channeling
the
synthesis of
possibility
recent physiological
in
the
both
vestibular
seems
evidence
nuclei
- 32
(Keller
and
Precht,
1978,
-
1979a,b; Waespe and Henn, 1977;
Daunton et al., 1979; Henn et al.,
1974;
Dichgans
et
al.,
1973; Allum et al., 1976) and behavioral experiments (Nashner
and Berthoz, 1978; Lestienne et
al.,
1977;
Vidal
et
al.,
1979).
The
visual
Keller
influence
and
on
vestibular nuclei.
semicircular
Precht
the
studies
firing
revealed
rates
ubiquitous
of neurones in the
Cells receiving input from the horizontal
canals
also
showed
.substantial modulation of
firing rate with horizontal rotation of the visual
such
that
surround,
rotation of the visual field had a similar effect
to rotation of the animal in
the
opposite
direction.
The
response dynamics of these cells were linear for combinations
of horizontal visual and vestibular stimulation;
the
rolloff
frequency for the response to visual field velocity was about
0.25 Hz.
Contrary to previous theories, the cerebellum was shown
to
be
unnecessary
for mediating these visual influences on
vestibular nuclei units.
not
show
adaptation
of
However, cerebellectomized cats
the
vestibulo-ocular
reflex when
reversing prisms are worn, indicating that the cerebellum
necessary
for
mediating such adaptation.
typically
saturate
at
visual
is
Saturation of the
response of the vestibular neurons occurred for visual
velocities above 3 to 10 degrees/sec,
do
field
whereas floccular units
velocities
greater
than
33
-
Though
1 degree/sec.
latter figure for the floccular
this
were
units may be of questionable accuracy since the results
obtained with anesthetized animals (Simpson and Alley, 1974),
a significant difference between the saturation velocity
and
floccular
further
units
vestibular
for
suggests
that
visual
input
non-cerebellar sources are responsible for the
to the vestibular nuclei.
One
for
method
in
unexpectedly
a
this
testing
is
drop
to
patterned chamber which can be
visually
dropped with the subject or moved opposite to
direction
the
In addition, the test should be performed with eyes
of fall.
closed and in the dark
field
visual
subjects
with
is
motion
necessarily
not
nulling
since
open,
eyes
visual input (Huang and Young, 1980; Waespe
of
equivalent to no
1980).
al.,
et
Any differences might be observed as changes in the magnitude
of
integrated
the
different
EMG
soleus
accelerations
have
"startle"
been
shown
different response magnitudes (Greenwood
Since
to
result
and Hopkins,
in
1976b).
visual cues in these experiments may be misleading, an
ideal strategy might be to minimize the influence
information
throughout
the
entire
testing
realizing that the cues are misleading.
these
since
reaction,
visual
of
sequence after
This
implies
that
experiments may demonstrate the minimum influence that
the visual input can have on the motor response system.
A completely
analogous
argument
applies
to
tactile
- 34
cues;
-
and other proprioceptive cues
tactile
unfortunately,
cannot be totally separated from
vestibular
cues.
In
account for all possibilities, the paradigm can be
to
order
the
as
considered
test
a
which
proprioception,
of
of
utilization
generalized
sensory input from tactile,
includes
receptors.
vestibular
and
muscle, joint, internal visceral
In the case of the early response, specifically, the reaction
can be considered as a
proprioception
generalized
intact labyrinths to provide the predominant
requires
which
of
test
component.
(1979) recently completed a set of animal
Vidal et al.
Baboons
experiments somewhat similar to those reported here.
were seated
in
chair
a
unexpectedly
was
which
Both this feature
Contact was slowed with an elastic system.
obviated
a
seated
position
pre-contact response in the leg muscles,
late
the
in
were
baboons
and the fact that the
dropped.
which rapidly
disappeared
short-latency
early
response
However,
repetition.
with
not disappear, and it is
did
this EMG reaction which was measured in
the
muscle.
soleus
early response was shown to be reduced in situations of
This
stabilized vision and darkness relative to falls in a
visual
environment.
to
baboon's head
normal
The
the
visual
reported
possibility
A
provide
lighted
box
stabilized
was
placed
vision,
normal
over the
whereas
the
field consisted of the laboratory environment.
results
that
the
might
baboon
be
explained
could
simply
by
the
preset the reflex gain
35
-
before
the
fall
beforehand.
depending
These
-
upon
issues
will
the
visual
be
discussed
early
response
conditions
further
in
Chapter 5.
Vidal
subdivided
the
subcomponents which could be readily distinguished
of the raw EMG.
from
In the baboon,
the first
two
into
in records
subcomponent occurs
60 to 100 msec after initiation of the drop.
Different
visual field conditions were reported to affect predominantly
the
second subcomponent, although the first subcomponent was
affected in certain
reported
that
cases,
sensory
subcomponent.
conditions
Lacour
labyrinthectomy
subcomponent, but only
other
also.
reduces
modalities
et
al.
eliminates
the
second,
contribute
(1978)
the
first
implying
that
to
the
second
However, Vidal has reported that visual
also
field
sometimes modify the amplitude of the first
subcomponent.
Nashner and
interaction
Berthoz
during
(1978)
rapid
tested
postural
reactions
antero-posterior sway during standing.
rails
could
be
suddenly
subject who was standing
surrounding
the
displaced
on
it
conditions
A
cart
to
induced
mounted
on
backwards, causing the
sway
forward.
A
box
subject provided a visual field which could
be moved relative to the subject
Three
to
visual-vestibular
were
during
investigated.
the
induced
sway.
In the first case (N,
normal) the box and cart were rigidly connected, resulting in
- 36
-
visual field displacement in the opposite direction
relative
to the direction
the
stabilized)
of
In
sway.
box was stabilized with respect to the head
so that no relative visual field
third
(E,
case
(S,
condition
second
the
motion
In
occurred.
the
relative visual field motion was
enhanced)
enhanced by reversing the box motion relative to the head; in
this
situation
the relative visual field velocity was twice
that of the normal (N)
condition.
Gastrocnemius EMG recordings from the above experiments
visual
revealed
100 msec.
onset
During the
to
100
the
on
influences
150 msec
early as
as
response
following
interval
of the stimulus, EMG in the S condition was suppressed
relative to the N condition.
the E condition also suppressed the response in
that
result
Perhaps more surprising was the
this time interval, but not as
These
much.
were
findings
as suggesting that any visual discongruence with
interpreted
other senses caused the short latency response to be reduced,
allowing more time for the discongruence to be resolved prior
to longer latency reactions.
the
of
responses
However,
gastrocnemius
increased magnitude of sway in the S
condition
This
clearly
longer
latency
resulted
condition,
and
in
the
an
E
reduced the magnitude relative to the N condition.
that
suggests
compensatory
are not.
the
in
the
nature,
longer
latency
reactions
are
whereas the short latency responses
- 37
Sequential
the
repetition of the S conditions revealed that
short latency (100 to 150 msec) response was susceptible
to adaptation; the reaction to
became
increasingly
the
similar
to
discongruent
conditions
that during the normal (N)
condition after several repetitions.
Thus,
Nashner
and
Berthoz
have
convincingly
demonstrated visual motion influence on rapid motor reactions
during a postural disturbance.
of
30
Assuming a transmission
to 50 msec from brain to gastrocnemius muscle,
information must be processed within the first
time
visual
70 msec
50 to
following onset of the stimulus.
Pathways involving the superior
suggested
possible
as
influences.
Maeda et al.
stimulation
of
routes
for
these
been
visual
fast
and vestibular nerve produce a
complex pattern of IPSP's and EPSP's in spinal
motoneurones,
is consistent with the fact that these pathways to the
which
revealed
that
forelimb
studies
Interaction
disynaptic.
motoneurones are at least
and
have
(1977) have shown in the frog that
tectum
both
colliculus
the vestibular and optic pathways to the neck
motoneurones
are
independent
and
not
do
significantly interact until the influences of these pathways
are integrated
should
at
obviously
visual-vestibular
the
not
motoneurone
be
level.
interpreted
interactions
are
to
this
However,
mean
transmitted
that
via
no
these
pathways, since both the tectum and vestibular nuclei receive
38
-
visual-vestibular
inputs.
visual
both vestibular and
It
is
possible
that
of a different nature can be
interactions
conveyed to the spinal level via both the vestibulospinal and
the
tectospinal pathways.
et
(Anderson
superior
1971)
al.,
Studies were performed in the cat
to
determine
influence
the
of
on neck motoneurones; it was determined
colliculus
that both tectospinal and tectoreticulospinal pathways may be
of
collicular
For a general overview of postural tests of
vestibular
responsible
for
mediating
the
effects
stimulation on neck motoneurones.
function, see
Appendix A beginning on page 190.
Anatomy and
physiology of brainstem structures relevant to the vestibular
system
and
posture
beginning on page 200.
control
are
reviewed
in
Appendix B
-
39
-
-
40
-
CHAPTER 2
DESCRIPTION OF APPARATUS
Because much of the equipment used in
custom
built
conditions, it
before
to
accommodate
is
detailing
appropriate
the
the
to
experimental
this
study
necessary
the
describe
was
stimulus
equipment
procedure itself.
This
chapter includes a description of the apparatus sufficient to
how
illustrate
the
proper stimulus conditions were created
and controlled and how data was processed and recorded.
The basic procedure required suspending a human subject
in
a
safety harness in an upright position and unexpectedly
releasing him upon deactivation
of
an
brake.
electric
A
checkered pattern covering most of the subject's visual field
Electromyographic
could be moved vertically during the fall.
(EMG)
in various leg muscles was recorded in order
activity
to observe the relationships between vertical
gravireceptive
and visual cues and the preparation for landing.
components
The major
of
the
apparatus
included
an
electric brake and controller; harness and suspension system,
plus
various
processor;
safety
devices;
and
accelerometer
visual
moving
field;
EMG
(for joint angle
goniometers
measurements); and PDP-11 computer, which output the stimulus
waveform
to
the
acceleration, and
brake
joint
and
controller
angle
data
to
converted
digital
form
EMG,
for
41
-
-
storage.
2.1
Electric brake and controller
steel
The harness was connected to a 1/8 inch diameter
which
cable,
passed
a pair of pulley wheels near the
over
ceiling and wound around a 4-inch diameter take-up drum.
figure 2.1
on
page 42.
The
See
take-up drum was mounted on a
shaft together with ,a Warner
electric
PB-500
See
brake.
Full activation of the brake armature
figure 2.2 on page 43.
to -90 VDC provided a static braking
of
torque
40 ft.-lbs.
is the equivalent of 240 lbs. of restraining force at a
This
drum radius of 2 inches, which was adequate for
all
the
of
subjects tested.
Maximum acceleration during a drop could be achieved by
the armature voltage to zero.
reducing
the
armature
and
take-up
achievable
acceleration
inertia,
drum
maximum
the
Any acceleration less
0.9 g.
was
Primarily because of
than 0.9 g could be obtained by reducing the magnitude of the
voltage
armature
90 VDC.
The
activation
90 VDC
from
torque-velocity
level
of
the
some value between 0 and
to
characteristics
brake
for
a
given
complicated the problem of
This
providing a constant restraining force during the drop.
was
partially
overcome
by
predetermining
the appropriate
armature voltage as a function of time following release.
The brake armature was powered
by
a
modified
Warner
-
42
(/
6stophits
spring when
subject
descends below
a safe level
vertically
cloth
pattern
right
side mirror
/1
electric
brake
knee angle limiter
platform for subject
to stand on, so that
U
cable can be reeled
in for next drop
athletic mat
0ee....000
e~.
.*** ...... * 0
0g
*eO
Se
..
.0a00000g..
eg
..
.
000aa00
e
*g*0
Figure 2.1.
Subject suspension system and safety
devices.
e
-
43
-
electromagnet and
friction face
brake
armature
steel cable
-
shock cord,
to maintain
tautness of ca ble
VOLTAGE-CONTROLLED
POWER SUPPLY
'1
I.E
6oHz
3/qA
N40y
max. excite/
external input
T P5..
switch
external
input
+
.
... ..
Figure 2.2.
O
q5K
f.zi
*
Braking system.
.
s
**0
*
*
06
-
44
-
The original device was only capable of manually
MCS-153-3.
controlled
on-off
switching
excitation
voltage.
This
and
had
potentiometer-adjustable
to
be modified to allow an
externally provided voltage from a computer
control
the
armature
voltage.
channel
to
The gain was adjusted such
that a 5-volt input resulted in the maximum
The
D/A
-90 VDC
output.
gain could be increased in order to accommodate subjects
weighing more than 160 lbs.
All except one
subject
weighed
slightly less than 160 lbs., and in these cases, weights were
added above the harness assembly in order to bring the
total
This precluded having to change the input
weight to 160 lbs.
waveforms for a desired acceleration profile.
In order to check the accuracy and repeatability of the
desired
acceleration
profiles,
the brake system was tested
The
with 160 lbs.
of iron weights attached to the harness.
repeatability
was such that time of.contact with the landing
surface varied
less
than 5%.
The
electromechanical
time
constant for armature deactivation was about 40 msec.
1 2.
Harness and suspension system
The safety harness
number 444)
was
(Mori Safety Products,
connected
to
the
figure 2.1 on page 42 and figure 2.3
tubing
guideways
prevented
the
Toronto,
part
steel cable as shown in
on
page 45.
subject
from
Aluminum
swinging
excessively and eliminated rotation about the vertical
axis.
The left and right steel cables to the harness passed through
+ + +
+ +
+ +
+++++++++
+ +
4-90
bO L
+ + + + +
s
+ + + +
-
+ + + + + + + + + + +
+ + +
++++++++++++++++++++++++++I
+
+ +
+ + + + + + + + + + + + + + + + + + + + + + + + + +
+
+
++
+ + + + +
+++++++++++++
+++
+++++++++++++++++++++++
++
++
+ +
++ +
++ + -+
+ +
+,
++++++++++++++++++++++++++
+++
+
+ +
+
+
.
+
+
+
+ + +
+
+
+++++
+
+A
+
+
+ + ++++
+ + + + + + + +
+ + + +
+ + + + + ++
+ + +
+
+++
+ +r-i-3+
+++m++++++
+
+
+
+ + + + + +
+ + + + +
+
+
+ +
+ + +
+ +
+ + + + +
+
+ + + + + + +
+++
+ + + + + +
+
+ + + + +++++ + +
+++
+
+
+ + + + + + + + + + + + +
+ + +
+ + +
++
+ ++ +
+ ++
++
++
+++++
++
+ ++
++
++
++
+
+r++
+
++ +
+ + +++ + +
+
+
+ + + +++
+ + + + +
+ + + +
+ '
+
+ ++++
+++++ + +
+
+ +
+
+ +
+ ++++++++
+ +
.
++
++
+
+
+
+ + + + + +
+ + + + + + + + + + +
+
+
+ +
a
5*
+ + + + + + + + + + + + + + + +
+ +
+
+
+ + + +
+ ++ +-++-
o
+
+ + +
eG+L++c++++++++++++
+
+
+
0+0++e+++++++ +
+
+ +
+
ri++o++++
+ + + + + + + + + + + +
+ + + +
+ + + +
+ + + +
++
++
+
+
+
+
+ + + + + + + + + + + +
+ + + + + + + + + + + +
+ + + + + + + + + + + +
+++++++++++++++++++
+ + + + +
+
+n
+ Q + + 0'+ + + + + + + +
+ + + + + + + +
+
+ +
+ + + + + + + +
+
+ +
+ +
+ + +
+ +
+ +
+ +
+ + O+
+ + +
+ + +
+ + +
++~I+++++++++++
-I
E1:
\
Iv-f
-+
.+
-r-
.
4-)
W-
-
-
-
-
-
O
.-0
-'0
-)-4-)
--
er-
- -IL+
e
*0)
0(00
bO
SC,).
-
46
-
the
the wooden 2x4, which slid freely on
the
Since
guideways.
holes in the 2x4 were oversized compared to the guideway
diameter, and the
was
friction
itself
guideway
insignificant
somewhat
was
flexible,
The
and jamming never occured.
large turnbuckles above the harness allowed the
assembly
to
be adjusted for subjects of varying heights.
case
As a preventive measure in the
landing,
inappropriate
the
collapsing to the floor by means of a
between
prevented
begin
to
spring
the
compress it whenever the subject fell below a
certain level.
This level was chosen to
the
fully
subject's
from
mounted
spring
large
The metal stopper mounted on the cable would hit
and
completely
pulleys wheels above the harness assembly.
two
the
was
subject
a
of
erect
In
height.
below
4 inches
be
a well controlled
landing, the maximum knee flexion was such that most subjects
did not fall below this level.
2.3
Additional safety measures
provide
Subjects were required to wear hiking boots to
ankle
support
and
to minimize the possibility of sprain or
injury due to accidental inversion
strap
when
the
slightly flexed.
than
foot.
the
A
nylon
was connected from the rear of the boot to the rear of
the harness, acting as a knee angle
that
of
subject
was
limiter.
suspended,
This
insured
his knees would be
The degree of flexion was not much
greater
would normally occur for a passive leg in the suspended
47
position, but did insure that active extension did not
before
contact
in
was
the
then
landing
a
the
preventing
all
Black no.
202495, medium; no.
subjects
Such
would
knees
harness
active
have
When the subject
became
unloaded,
Full knee extension
unnecessary
fatigue.
In
wore elastic knee braces (Bauer and
All landings were on
202657, large).
one-inch-thick
athletic
matting
type of surface, as well as boots and knee
The
supports, may significantly
affect
the
parameters
of
the
but since all of the tests were performed
postural responses,
these
the
straps to slacken.
nylon
addition,
using
of
locking
surface,
possible,
(Resilite).
surface.
a jolting and unsafe landing.
was standing on
allowing
the
possible
extension and
resulted
with
occur
this
mats,
factor
was
not
an
uncontrolled
variable.
2.4
Visual field motion
Because of
the
importance
of
peripheral
vision
in
detecting motion, the visual field motion system was designed
to allow as much field coverage as possible.
loop
of
cloth
was
stretched
taut
aluminum tubes, one near the floor and
ceiling,
12 feet
high.
tube could be connected to
assembly
via
A
7-foot
wide
between two horizontal
the
other
near
the
See figure 2.3 on page 45.
The top
the
harness
the
wooden 2x4
of
a 1/16-inch diameter steel cable on either the
left or right side of the tube.
Each end of the top tube was
-
48
-
fitted with an adapter with spiral grooves such that the left
and right cables
were
wound
in
opposite
directions.
By
choosing to connect either the left or the right cable to the
harness assembly, the visual field could be driven up or down
at
exactly
the
same speed as the subject.
attaching an extension cable to
cable,
the
visual
field
either
the
In addition, by
left
or
right
could be moved manually while the
subject remained suspended.
Acceleration of the subject during the fall was
less
than 3%
because
of
the
additional
slowed
inertial
and
frictional loads of the visual field motion system.
The cloth
pattern
alternating
one-inch
was
blue
blue
and
vertically and horizontally.
an
example
of
that
white
gingham
stripes,
the type of pattern used.
cloth
with
running
See figure 2.4 on page
the subject to the plane of the
Assuming
checkered
49
for
The distance from
pattern
was
76
cm.
the subject fixates on a point level with his
eyes, the spatial frequency varies from about 0.2 cycles
degree
at
the
from the fovea.
relative
per
fovea to 0.8 cycles per degree at 60 degrees
The
positional
pattern
cues
spatial frequencies in this
was
were
range
quite
uniform
unlikely.
are
quite
so
that
Patterns
with
effective
generating linear vection sensation (Lestienne et al.,
this particular pattern was chosen
powerful visual velocity cue.
in
order
to
in
1977);
provide
a
- 49
-
III!!
!!!!!!!
!!!
1!!! !!!
HI!!!
IHI
-- ########## ---------- #########
------
########## ---------- ##########
-------- ########## ---------- ##########
#----------##########----------##########
## ---------- ########## ---------- #########
##### ---------- ########## ---------- #########
II I !!
M M!!!!!
! !!
HIMM
!!!!!!
M I M!!!!!!!!!!!
I
######### ----------
M!!!! !
!M!! !
M
!
##########----------
! !! !!
########
######## ---------- ########## ---------- ########
######## ---------- ##########----------#######
######## ---------- ########## ---------- #######
######## ---------- ########## ---------- ######
######### ---------- ########## ---------- #####
IM!! ! !
M! M!!! M !
!!!
1
I !
!!! M
! !M ! !
M
!!!!!
1!
IM
I !!!!!
!!! M
M!!
!
!
M!!
IM
II!
####### ---------- ########## ---------- ####
##### ---------- ########## ---------- ###
# ---------- ########## ---------
Figure 2.4.
Example of type of visual; f ield
(scale, 1:1).
pattern used
- 50
A pair of 12-foot-high,
metal-backed
acrylic
mirrors
were placed at 60-degree angles to the cloth pattern in order
to extend reflections of the pattern
regions
the
of
visual
field.
into
The
noticeable imperfections in flatness
but
these
periphery;
the
acrylic
when
lateral-most
mirrors
viewed
had
directly,
defects were not objectionable when viewed in the
peripheral vision is very
sensitive
to
velocity
cues, but is notably insensitive to pattern variations.
Ski goggles with clear (non-tinted) lenses were worn by
subjects in order to eliminate areas of the floor and ceiling
Stationary regions of the peripheral visual field
from view.
tend
to
cancel the vection effects of moving regions, so it
was considered important to attempt to eliminate the
No
stationary
were visible using the apparatus and
regions
procedure outlined in this study.
for
a
plot
of
the
effect.
See figure 2.5 on
page 51
approximate region of the visual field
remaining with the goggles on, compared to
the
unrestricted
The actual area may have varied somewhat, but in all
field.
cases none of the subjects could see the
floor
when
facing
variability
in
effective field area arose for myopic subjects who needed
to
straight
wear
ahead.
corrective
the
Perhaps
lenses.
The
greatest
goggles
fit over the lenses
without difficulty, but it is possible that myopic people who
wear
glasses
may
interpret
corrected field differently.
visual
influences outside the
-
51
scales marked
in de gre es
90
30
-~?
-CC
s3'0
40
30
9.0
I.
RIGHT EYE
LEFT EYE
boundary of
vision with
goggles on
Figure 2.5.
boundary of
normal vision
Map of restricted visual field using
goggles.
- 52
The
visual
field
light,
incandescent
The
lighting.
was
illuminated
order
spatial
with
with
effects
the pattern was high
of
frequency
fluorescent
and
enough for the latter problem to be of concern,
provide
would
undesirable
indirect
eliminate shadows and to
to
stroboscopic
possible
prevent
in
-
stationary cues.
incandescent lamps were aimed at the walls to
shadows
Eight 150-watt
left
the
with front and rear shields to eliminate direct light
right,
Shadows
and to minimize light reflecting from the rear wall.
on
and
the
cloth
pattern
almost
were
imperceptible and very
diffuse.
2.5
EMG electrodes and preamplification
Differential EMG signals were obtained from each muscle
of
interest
a
using
Hewlett-Packard pre-gelled,
of
pair
disposable, Ag/AgCl surface electrodes
with
snap-on
type
electrode
leads,
(part
number 14245B,
number 14279A).
part
These electrodes have a wide adhesive collar which
was
very
effective in preventing accidental detachment during testing.
The
electrode
differential
pairs
preamplifiers
The primary component of
Devices
provided
which
each
the
of
to
small
were strapped to the leg.
was
preamplifier
an
Analog
AD521K instrumentation amplifier with a CMRR greater
than 90 dB and used with a circuit gain of
pair
input
100 kilohm
input in order to
100.
A
matched
resistors was placed in series with the
decrease
the
possibility
of
accidental
- 53
electric
The
electrodes
and
tapping
See figure 2.6 on page 54 for a schematic.
shock.
were
pushing
initially
on
checked
them.
No
for
motion
artifacts
by
artifacts were
visible when using the application procedure described in the
next chapter.
Differential
eliminated
preamplification near the source virtually
electrostatic
and
electromagnetic
normally caused by long leads from
However,
fibre
impedance
sources.
another source of distortion is introduced which is
due to the nature
involves
high
interference
the
of
effect
depolarization
the
differential
configuration,
and
of the travelling wave delay of muscle
between
the
two
electrodes
(Hogan,
1976).
2.6
EMG processing
a
Each channel of the EMG processor consisted of
stage,
band-pass
stage,
gain
full wave rectifier, and averager.
Gain could be varied from 0 to 10,
and the raw EMG
could
be
recorded directly from the output of this initial gain stage.
20
Band-passing was adjusted to include frequencies from
1000 Hz.
to
This range of frequencies has been demonstrated to
be highly correlated to muscle force
1 KHz, the correlation decreases.
(Hogan,
Above
1976).
Full wave rectification is
one of the better methods for obtaining
force
a
signal
(Kreifeldt
which
and Yao,
highly
correlated
1974).
The rectified signal was then passed through a
with
muscle
is
third
54
-
-
100
1(,0k/B
gain=
to electrodes
OP
100
io
A '5 1
to EMG
processor
(--I
Fgure 2.6
0
ge
*
.06
000
Elcrd
E
e
ct
ode.
.
preamplifier schematic.
.
.
...
.
.
.
.
sch*i
.
0
*
@
O
eeaeea
e
0rr
- 55
order
averaging
filter
(Garland
et
increases the signal to ripple ratio
al.,
by 50%
1972),
which
to
compared
pole RC filter with a comparable step rise time.
single
was
averaging time during all experiments
to
set
a
The
10 msec.
The averaging time is defined to be the time required for the
Thus, this
step response to reach 95% of steady state value.
time would contribute only several msec compared to the brake
time constant for the purpose of determining
EMG
latencies.
See Appendix C for a circuit schematic and transfer functions
of each stage.
2J7
Acceleration measurements
A
single
(Statham
axis,
strain-gauge
F-10-350,
Laboratories
bridge
-10
to
accelerometer
+10g
range)
was
mounted to the cable above the harness assembly to verify the
actual acceleration profile during each drop.
for
a
schematic
of
During
tests
in
visual
field
pattern
accelerometer
was
the
accelerometer
was
preamplifier.
subject was stationary and the
the
which
and
See figure 2.7
manually,
moved
to
attached
an
identical
the extension cable used to
pull the visual field pattern.
2-,_8
Goniometers
Knee and ankle flexion measurements were
prototype
Arthritic
goniometer
and
system
Rheumatism
developed
Society
and
by
the
made
the
with
a
Canadian
University
of
56
-
3MSe
CC
*O.
*
4
00
A-t
Figure 2.7.
Accelerometer and amplifier schematic.
- 57
The knee and
1976).
and Cousins,
British Columbia (McKechnie
axes
flexion
the
of
the
and
ankle goniometers were connected by a sliding link,
were
potentiometers
rotation
The entire assembly did not noticeably impede
self-aligning.
movement, and the accuracy in this particular application was
estimated to be better than 3%,
and
testing
during
slippage
of
based on observed amounts
manufacturer's
of
claims
See figure 2.8 on page 58 for a diagram of
1% accuracy.
the
goniometer assembly.
The goniometer system was applied to the left leg only,
the right leg being completely occupied by EMG electrodes.
harness
hip goniometer section was available, but the safety
the
holding
subject
made
goniometer impracticable.
has
communication)
are
well
that the
indicated
and
timing
in
both
feet
make
reactions
the
that
two
hip
(personal
Watt
milliseconds of each other.
Computer control of stimuli and data conversion
Initiation of the fall and data conversion
the
the
of
While it was possible that
coordinated
contact within a few
2.9-
placement
occurred between the two legs,
differences
legs
secure
A
control
of
a
PDP-11/34
computer.
were
under
The waveforms for
controlling the brake activation during the fall were
stored
on disk and transferred to memory prior to output via the D/A
channels.
of
an
LED
The only other real time output was the activation
in
the
experiment
room
to indicate that data
-
58
-
flexible
parallelogram
chain
knee goniometer
signal
knee flexion
potentiometer
ankle goniometer
signal
sliding
link
coupling
flexible
parallelogram
chain
ankle flexion
potentiometer
Figure 2.8.
Leg goniometer assembly.
- 59
begun;
had
conversion
experimenter
was
this
used
as
a
cue
the
by
to begin moving the visual field in cases where
the subject remained suspended during visual field motion.
Eight A/D channels were available; five
convert
were
used
to
filtered EMG output from the EMG processor, two were
used for the knee and ankle flexion signals, and one was used
to
convert
the
amplified
acceleration
signal.
This last
channel was connected to either the amplified signal from the
accelerom'eter
above
the
harness,
accelerometer connected to the
visual
field
operation.
or
the
extension
(See
cable
section 2.3
visual field system.) The sampling rate was
channels.
signal from the
for
manual
describing the
500 Hz
for
all
One second of data was taken from each channel and
stored on disk.
2.10
Miscellaneous equipment
Each subject wore a crash helmet to provide
against
potential
falling
convenient anchor on
masking.,
level
equipment noises.
prevent
to
place
and
to
provide
earphones
for
noise
to each drop
to
distract
from
It was switched
and switched off
habituation (and
most,
but
not
all,
on at a random time just
immediately
possible irritability).
after
metal stop hitting
the
spring
above
the
to
The only
loud, impulsive noise occurred after the subject landed,
to the
a
The masking noise consisted of Gaussian noise at a
sufficient
prior
which
objects,
protection
due
harness
-
assembly.
60
(See figure 2.1 on page 42.)
- 61
-
-
62
-
CHAPTER 3
EXPERIMENTAL PROCEDURE AND
STATISTICAL DESIGN
This chapter outlines the actual procedure followed
prepare
the
install him in the apparatus, and
and
subject
perform a sequence of drop tests.
discussion
of
the
overall
This
is
followed
design
experimental
by
a
from
a
For a more detailed description
statistical perspective.
to
of
the rationale behind the use of a particular device, refer to
the previous chapter.
3,_.1
Preparation of the subject
All
orthopedic
subjects
were
and
neurological
by a surgeon specialized in posture and
problems
gait disorders.
for
examined
Prior
history
of
broken
bones,
sprained
ankles or knee joints, stretched ligaments, backache, or poor
reflexes was sufficient to disqualify a potential subject.
vestibular
system
which included
horizontal
diagnostic
checks
rotation,
for
procedure was also performed,
abnormal
positional
eye
tests
of
semicircular
canal
movements
nystagmus,
response to an air caloric vestibular test.
all
affected.
Currently,
tests of otolith function.
and
during
abnormal
While these
are
function rather than the
otoliths, in most cases of otolith disorders the
also
A
canals
are
there are no selective clinical
-
63
Each subject witnessed a demonstration drop test before
participating.
A brief questionaire was given to the subject
which included subjective estimation
anxiety
and
both
alertness,
of
own
his
level
of
on a scale from 1 to 7, and a
report of past sports activities.
All of these factors might
be expected to have an effect on the patterns observed in the
EMG responses.
Prior to electrode'application, each site
wiped with alcohol,
demonstrated
to
consistently
paths from electrode to muscle and
This method
provide
the
skin;
reducing
artifacts
A large component of
motion artifact is due to variations in skin
on
the
skin
potential
reflected in the measured voltage.
with
resistance
scratching diminishes the degree to which the skin
is
has
low resistence
muscle
reduce
(Tam and Webster, 1977; Webster, 1977).
pressure
shaved,
and scratched with a hypodermic needle in
order to break through the stratum corneum.
been
was
by
potential
Raw EMG (preamplified
but unfiltered) was recorded with an FM tape recorder
during
experiments on the first three subjects to check for possible
artifacts.
motion
observed
occurred
from contact with
swings
which
were
were
ever
when an electrode had accidentally
broken
The
the
only
skin,
also
which
artifacts
creating
very
large
voltage
quite obvious in the filtered EMG.
Because of this, it was judged unnecessary to record raw ENG.
During
the
testing,
EMG
signals
were
recorded
-
simultaneously
from
the
-
quadriceps
femoris),
(biceps
hamstrings
64
(rectus
femoris),
anterior,
tibialis
gastrocnemius and soleus muscles of the right leg.
right
Electrode
pairs were placed over the approximate middle of each
and
body,
single common ground electrode was placed over
a
See
apart.
5.7 cm
spaced
electrodes were
oriented
in
the
the muscle, and the centers of the
of
direction
was
Each pair
ridge.
the anterior tibial
lengthwise
muscle
figure 3.1
on
page 65 for exact electrode locations on the right leg.
the subject then put on
After electrode placement,
elastic
knee
braces,
and the knee and ankle
right leg and the
The
connected.
EMG
goniometer
goniometer
and
while asking the subject to
the
to
attached
were
preamlifiers
subject's
boots,
In the experiment room the input leads
goniometers.
EMG
hiking
hop
up
the
to
the
electrodes on the
cables
signal
were
channels were checked
and
Then,
down.
the
subject put on the crash helmet and goggles and was placed in
Knee angle limiters were attached from the rear
the harness.
of each boot to the harness and were adjusted.
3.2
Basic procedure far each drop
At this point
procedure
for
the
subject
the first drop.
was
ready
to
in manually to a predetermined
turnbuckles
above the
harness
the
First, a platform was placed
under the subject so that the cable to the harness
reeled
begin
were
could
calibration mark.
adjusted
so
that
be
The
the
- 65
-
rectu sS
f emor is
biceps
femoris
right head
of
gastrocnemius
groun d
tibia lis
anter ior
soleus
REAR OF
RIGHT
LEG
FRONT OF
RIGHT LEG
0
Figure 3.1.
Leg electrode positions.
0 a*00
00
a
- 66
adjustment was necessary only
before
the
the fall began, the harness position was
the
held
by
and
Until
switching
The platform
controller to "maximum excitation".
brake
drop,
first
height for all drops was the same.
the
that
insured
This
barely resting on the platform.
were
heels
subject's
a
height could be chosen to be either 28 or 43 cm for
given
test.
With the subject still standing on the platform, he was
to close his eyes.
told
The masking noise was then switched
on and desired mechanical connections
motion
were
system
made.
to
in order to
prevent
subject from deducing the nature of the subsequent test.
After this procedure, the subject was
close
field
visual
This procedure was followed even
when no visual field motion was desired,
the
the
instructed
to
either
or open his eyes, depending upon the parameters of the
test.
To standardize the initial head position,
was
eyes
were
In cases in which
closed or the room was darkened, the subject
was asked to position his head and eyes in the
as if he were looking at the screen.
out from beneath, and the drop
but
subject
also instructed to maintain the head erect and fixate on
a point on the screen level with the eyes.
the
the
within
the
next
then
10 seconds.
same
manner,
The platform was pulled
occurred
Although
extension of the head was not measured, no
head
unexpectedly,
flexion
and
motion
was
-
during
visible
the
Since the head was constrained
falls.
by
from further extension
the
of maximal extension allowed by the
and
and
helmet
the
suspension
either side, the subject's head was in a position
to
cables
67
during
the
fall.
apparatus,
both
before
This constraint was such as to allow
the head to be maintained in a
erect
position.
After
the
fall, the subject was told to switch the masking noise off.
To summarize the possible test conditions:
of
the
drop
the
height
could be 28 or 43 cm; the acceleration profile
could be selected from a
pre-synthesized
set
of
profiles,
including the case when no drop occurs; and, the visual field
could move down with
the
subject,
it
could
move
exactly
opposite to the subject's motion, it could remain stationary,
or it could be moved manually.
the
dark
(eyes
open)
The drop could also occur
or with eyes closed.
in
Thus, the only
variables of which the subject was aware beforehand
was
the
initial height, and whether the room was darkened or his eyes
were closed.
This entire procedure was followed for all experiments,
except
for minor variations in certain cases.
In the report
of the results, any such exceptions to this procedure will be
mentioned.
3.3
Statistical design of experiments
The
investigation
of
the
primary
effects
of
- 68
profile,
acceleration
-
initial
height,
conditions is complicated by possible
as
occur
the
experiment
proceeds
visual
and
adaptation
and
by
field
which
may
possible
the
interaction of previous tests with the reaction in subsequent
example,
For
tests.
misleading cue,
influence
of
since a moving visual field provides a
vision
on
immediately following.
were
of
prime
reduce
the subject might be expected to
reaction
his
Therefore,
statistical
the
test
considerations
determining the experimental
in
importance
during
the
design and the order of testing.
A three way split plot design was chosen
as
a
method
which would allow separation of the effects of adaptation and
interaction of adjacent tests as well as the main effects
and Cox, 1950).
In a three way split plot design, there
R(rows)
x
B(blocks)
x
represents a different subject,
particular order of testing.
used
in
each
block.
position
in
Each
the
a
drop,
each
and a block represents a
A different set of subjects was
repetition
conditions must be categorized in a
temporal
are
C(columns).cells, where each column
represents a particular set of conditions for
row
1969; Cochran
and visual field motion (McNemar,
acceleration
of
testing
new
of
a
column,
sequence
set of drop
since
must
the
also be
considered as a variation of the drop conditions.
See
statistical
Appendix D
for
significance
the
for
procedure
for
determining
main block and column effects
-
69
-
and block by column interactions.
Actual combinations
field
parameters,
for each
chapters.
set
of
of
acceleration,
height,
visual
and the order of testing will be specified
experiments
discussed
in
the
following
- 70
-
- 71
CHAPTER 4
EXPERIMENTS
AND RESULTS
The experimental procedures described in
chapter
were
between the ages
One
females.
on
completed
subject
a
had
scoliosis, which did not pose
case
of
were
whom
of asymptomatic lumbar
was
but
risk,
safety
a
thirteen subjects
two
30 years,
and
20
of
of
total
a
previous
the
of
concern
because scoliosis may affect reflex characteristics.
However,
no unusual variations
observed
were
in
the
data
See table 4.1 on page 72 for
relative to the other subjects.
a brief description of the ten subjects who were included
the final,
perfected
paradigm.
Vestibular system function was tested for normality
a
in
by
series of standard clinical procedures which are mentioned
No abnormalities were indicated
in the previous chapter.
in
any of the subjects.
4.1
Stimulus combinations
acceleration
A total of five different
tested,
which
are
shown
in
figure 4.1
also
shown.
These
each
for
corresponding velocity profiles
profiles
were
page 73.
The
acceleration
are
on
included a 0.5g step, a 0.85g step, two
profiles with a double acceleration step, and an acceleration
ramp.
An
additional
condition
subject was not dropped while
the
was
included in which the
visual
field
was
moved
-
Table 4.1.
72
-
Description of subjects.
subject I5
initials
sex
height
(ft,in.)
weight
(lbs.)
Block 1:
EMA
M
5'10"
145
PNE
M
6' O"f
160
PRI
F
5' 8"
150
moderate activity in skiing
myopic, wore glasses during
testing
DDL
M
6' 1"
155
weight lifting,
TED
M
51 91"
137
tennis, basketball, soccer
M
5' 7"1
140
some swimming; past skydiving,
but not currently active
RWA
M
6' 0"
175
running,skiing, Zen meditation
PPE
M
5? 7?!
145
some swimming and
weight lifting
myopic, wore glasses during
testing
LLA
F
5? 5"
111
considerable activity in
gymnastics
APA
M
5t 9"?
115
fencing, Tae-kwon-do
myopic, wore contact lenses
during testing
Block 2:
TWA
sports activities, misc.
moderate activity
in skiing
none
bowling
- 73
-
velocity at contact
2.2 m/sec
0.5G
,cel*
2.6 m/sec
0 C*
0.85G
2 .2 m/sec
SDG
7~
44e $~
- --- V
V
IV
2.3 m/sec
FDG
qe
2 .2
m/sec
RG
eel
Figure 4.1.
Acceleration
and velocity profiles for
acceleration stimuli used in
drop tests.
100 msec
19
1.4
m/sec
- 74
In
unexpectedly.
to
order
-
facilitate
conditions will be. referred to
double
step),
0.5G,
as
discussion,
these
SDG
(slow
0.85G,
and OG (no
FDG (fast double step), RG (ramp),
drop).
as
Visual field conditions were
visual
effects,
field moved up while
during
closed
moved
field
the
the
fall,
darkened during the fall.
to
and,
was
nothing
in
the
dropped,
eyes
were
eyes were open but room was
referred
These conditions will be
as N (normal), D (down), U (up),
(dark), respectively.
during the drop, visual
down
subject
unusual
no
follows:
EC (eyes closed), and DK
Subjects reported that they could
darkened
see
Only 10 seconds were
condition.
available for any dark adaptation, which was insignificant.
and
eyes
Since the distance, d, between the subject's
the screen was 76 cm, the relative angular visual velocity at
the
by
the central part of the visual field -can be obtained
following relationship:
'0 for /D
n =1
0)V1s =Y
condition
for /N
2 for /U
where
6y~s
is the relative visual field velocity
the actual downward velocity of the subject.
condition,
in
m/sec
(as
is
For example, to
obtain the relative angular velocity in degrees/sec
IN
v
and
for
the
multiply the subject's actual linear velocity
shown
in
figure 4.1)
by
the
factor,
- 75
-
-75 degrees/m (-1/0.76m x 57.3deg/radian).
A
total
acceleration
of
fifteen
different
combinations
of
and visual field conditions were tested and are
listed in table 4.2
on
page 76.
The
acceleration
double
steps and ramp cases were tested only under the normal visual
these
field condition, since the primary objective of
was
to
determine
the
tests
ability of the subject to respond to
acceleration changes after the initial release.
4.2
Preliminary tests
Several modifications
were
made
of
the
experimental
procedure
after testing the first three subjects.
were
of problems
encountered
during
these
A number
initial
tests
concerning issues which are relevant to the interpretation of
the data.
For the first three subjects, the design of
drum
axis.
This
resulted
profile during unwinding.
this
brake
was such that the cable would sometimes wind around the
drum in a plane which was not exactly
drum
the
problem
was
in
variations
By adding
eliminated
and
to
perpendicular
grooves
the
of acceleration
to
the
drum,
the acceleration profile
achieved a much greater degree of standardization.
Acceleration measurements on the first
were
made
by
three
subjects
fixing one accelerometer to the safety helmet
and another to the subject's boots.
Due
to
variations
in
- 76
Table 4.2.
Combinations of acceleration profile and
visual field conditions which were tested.
acceleration profile
O.5G
visual field condition
N (normal)
D (field moved down during fall)
U (field moved up during fall)
EC (eyes
closed)
DK (darkened
0.85G
N
D
U
EC
DK
SDG
(slow double step )
N
FDG
(fast double step )
N
RG (ramp)
N
OG (no fall)
D
U
room)
- 77
of
monitoring the accuracy
as
unacceptable
was
this
position,
head
acceleration
the
method
for
profile.
An
a
accelerometer was placed on the boot in the hope of obtaining
did
signal
a simple indicator of moment of contact, but the
not allow a precise determination of contact; plantar flexion
of the foot usually cushioned
to
sufficiently
landing
the
eliminate sudden deceleration.
The acceleration measurement repeatability was improved
by
attaching an accelerometer on the cable above the harness
page 45.)
on
figure 2.3
2x4
the
above
assembly
this placement was that vibration artifacts
of
superimposed
vibration, though.
Acceleration measurements recorded during
figure 4.2
on
acceleration measurements
during
page 78.
at
the
unreel
160-lb.
Compare
cable
load
are
to
the
these
which
were
taken
Since the cable
continued
a short distance after the subject contacted
for
the floor, the
irrelevant.
a
drop test; see figures 4.3.1-14 on pages 85-98,
each
the sixth channel of each figure.
to
the
of
most
dampened
calibration tests at the harness with
in
structure
support
on the acceleration signal as measured at
the harness; the harness effectively
shown
major
The
vertical.
arising from the natural frequency of the
were
(See
pulley.
This allowed accurate placement of
the accelerometer with respect to the
disadvantage
the
beneath
but
acceleration
The
main
information
utility
after
contact
is
of acceleration measurement
during a drop was as an indicator of release time
and
as
a
- 78
-
0. 5G
_.85G
SDG
FDG
RG
1g
1100 msec
Figure 4.2.
Acceleration measurements taken at the
harness with 160 lbs. of added weight for
calibration.
- 79
-
gross check of repeatability.
as
not
was
as
great
treatment
statistical
acceleration
any
data,
the
of
of the
because
However,
desired.
profiles
acceleration
In general, the accuracy of the
variability would be incorporated into experimental error and
the
in
correctly reflected
values
significance
obtained.
test conditions were compared within the data for each
Since
subject, any consistent bias
subject
factored
be
would
given
a
for
acceleration
in
The main source of such a
out.
bias would possibly be failure to accurately
compensate
for
with
the
weight differences among subjects.
The goniometers which were available for use
last
greatly
subjects
ten
dorsiflexion
of
the
sudden
a
the ball of the foot hits the ground.
as
records
goniometer
intermittent
-determination of
undergo
to
forced
is
contact time; ankle angle
the
improved
with
connections
jolts during landing.
manifested
the
artifacts
power
However, the initial
due
Some
to
supply caused by
onset
of
ankle
dorsiflexion always preceded any such artifacts, and provided
a reliable indicator of contact.
In addition to the stimulus measurement
procedure
more
used
frequent
the
in testing the first three subjects revealed
fundamental
responses.
problems,
Since
difficulties
the
adjustments
procedure
of
the
obtaining
in
had
repeatable
just been developed,
apparatus
necessitated
-
interruptions
80
of the experimental sequence.
By the time the
fourth subject was tested, the intertrial interval was
never
more than several minutes, which probably reduced fatigue and
boredom significantly.
For the
listed
three
first
subjects,
each
test
condition
in table 4.2 was repeated at heights of 28 and 43 cm,
three times at each height.
required
This
a
total
order
Since testing order was one
have been excessively fatiguing.
the
in
Longer times would
to complete the session within two hours.
of
90
These 90
trials, randomized with respect to test conditions.
trials had to be broken down into several sessions
of
important conditions of the experiment, breaking the
sequence into several sessions made
effects
sequencing
single session.
session,
as
impossible
except
addition,
failure
exactly
on
electrodes
In
well
as
the
same
changes
in
of
interpretations
tests
within a
replace
surface
for
to
location
the
the
from session to
subject's
physical
condition, would make comparison even more difficult.
Since the subject could observe his initial height
could
successfully account for this variable,
28 cm were eliminated from the
subjects.
sequence
for
and
the drops from
the
last
ten
One of the primary objectives of this study was to
investigate ability to respond to cues during the fall, so it
was
considered
desirable to standardize conditions prior to
release as much as possible.
-
Thus,
with
the
45 -trials
single
in
the
ten subjects.
Any
The procedure as described
the
to
applies
chapter
and the
a
during
the
was considered adequate for testing
procedure
without further changes.
previous
of
standardization
the addition of goniometers,
shortening of the procedure to
session,
-
improved
the
profiles,
acceleration
81
last
which are given in the text use only the data from
statistics
these last ten subjects.
1L.3
Testing order and coding .2f. trials
plot
Since the split
effects
block
main
to
sensitive
experimental
of
interactions when the number
blocks
were
number
of
available
considered
is
is
highly
the
dividing
to
pool into a larger number of blocks;
subject
two
large,
of the limited
Because
superior
is
by treatment
block
treatments
adequate.
this
subjects,
and
design
data
from a large number of blocks with few subjects in each block
be
would
value, since the likelihood that block
lesser
of
differences were due to real interblock
subject
differences
would increase.
In this experiment each of the two blocks represents
different
order
of
testing.
exact reverse of the other.
preceded
simple
effects
by
a
Each testing sequence is the
This
allows
each
test
to
be
a different test in each block while permitting
comparisons
because
of
overall
of the symmetry.
adaptation
or
habituation
Testing sequence for both
-
blocks is
page 84
shown
in
82
table 4.3
on
page 83.
Table 4.4
on
groups the tests by stimulus condition and lists the
sequence position for each test.
In order to facilitate references
tests,
to
particular. drop
a code will be used in the discussion which is of the
form acceleration/
visual
repetition number.
For example, the code O.5G/EC/B2/3 refers
field
condition/
block
number/
to the third repetition in block 2 of the test for which
acceleration
was
0.5g
and
the
subject's eyes were closed
the
during the fall; the code O.5G//B1/ refers to all 0.5G
drops
in block 1.
4_,4
General d'escription _of .the EMGQ response
Several examples of actual EMG recordings are. shown
figures 4.3.1-14 on
many of the
section.
pages 85-98.
characteristics
Each
page
shows
three repetitions for a given
in
These examples illustrate
which
are
discussed
superimposed
subject
and
in
this
waveforms for all
test
condition.
The waveforms were reconstructed from the sampled versions of
the original raw data; thus, the top five waveforms represent
the sampled, rectified, filtered EMG data.
In the discussion which
follows,
the
five
pairs
of
electrode leads will be referred to as GN (gastrocnemius), SL
(soleus), TA (tibialis anterior),
(biceps
femoris).
These
RF (rectus femoris), and BF
muscle
abbreviations
refer
(cont. on p. 99)
-
Table 4.3.
83
Drop test sequence for block 1 and
block 2 (reverse of block 1).
test code
0.85G/U
OG/D
0.5G/EC
0.85G/EC
0.85G/DK
0.5G/D
0.85G/DK
0.5G/EC
0.5G/N
SDG/N
sequence number:
block 2
block 1
1
2
3
4
5
6
7
8
45
44
43
42
41
40
39
38
9
10
11
12
13
14
15
37
30
0.85G/DK
0.5G/DK
16
17
18
19
20
21
22
23
24
25
26
27
28
29
0.85G/D
30
0.85G/EC
31
32
0.85G/D
OG/U
0.5G/EC
0.85G/N
0.5G/D
0.85G/N
SDG/N
0.5G/N
0.5G/U
FDG/N
0.85G/U
RG/N
0.85G/EC
OG/D
0.5 G/N
0.85G/U
0.5G/D
0.5G/DK
FDG/N
RG/N
OG/U
0.5G/U
OG/D
OG/U
0.5G/U
FDG/N
0.85G/N
0.85G/D
SDG/N
0.5G/DK
RG/N
33
34
35
36
37
38
39
40
41
42
43
44
45
36
35
34
33
32
31
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
-
Table
84
Drop tests grouped by test type,
4.4.
giving sequence positions in blocks 1 and 2.
test code
BLOCK 1
sequence
position
0.5G/N//1
2
3
/D//1
2
3
/U//1
2
3
/EC//1
9
18
25
6
15
27
19
36
39
3
2
8
3
/DK//1
2
3
13
29
32
44
0.85G/N//1
2
1
/D//1
2
3
/U//1
2
3
/EC//1
2
3
/DK//1
14
16
L41
11
30
42
1
21
26
4
23
31
5
2
7
3
28
SDG/N//1
2
10
17
43
20
FDG/N//1
2
RG/N//1
2
3
0G/D//1
2
33
40
40
22
34
4q
2
24
mean
sequence
position
17
16
31
8
35
28
16
19
13
3
35
38
29
30
15
38
38
43
2
14
17
11
5
30
22
32
4
16
35
20
25
45
15
23
42
18
30
39
33
18
27
3
23
31
34
29
23
36
' 66
13
26
261
15
12
P4
12
22
25
214
9
21
44
12
2
21
28
37
19
31
40
7
10
27
33
me an
sequence
position
41
17
/U//1
BLOCK 2
sequence
position
28
8
11
314
18
-
85-
)100vSC
0S
L
T
A
0e
k1 I< t
00440
pla a 4r
dor~ilKO
Figure 4.3.1.
1exiov'
r tn.c~
+
Subject EMA, tests O.5G/N/B1/1-3.
- 86
-
loo risec
*ire,=
TA
k1.
S.............
Figure 4.3.2.
Subject EMA, tests 0.85G/N/B1/1-3.
87
-
10 wse c
I
I
Pr'
SL00
*
a
~ ~~~~
AjYtV~
_
4b
/J-cV-ecl
...
......
...........................
On r%1
7Figue
40.3.
Figure
4.3.3.
Subject EMA, tests RG/N/B1/1-3.
0.
- 88
-
tOo ~
Si..
TA
03
i
0
accel
-to
accelerometer connected
cable which moved
the visual field
kn-ee-
ari~.
Figure
4.3.4.
Subject EMA, tests OG/D/B1/1-3.
-
N-
89
-
t"
ce.
-.
Figure 4.3.5.
Subject DDL, tests O.5G/N/B1/1-3.
-
90
-
100mMNec
S'IN
acc__
arv
Figure 4.3.6.
Subject DDL, tests 0.85G/N/B1/1-3.
-
91
soO
eC
N Raw
Immomw
=
-
d%=
,-
SL-
_w=
:L, --
TA
C~t
1!
e
Figure 4.3.7.
Subject LLA, tests O.5G/N/B2/1-3.
....
.......
- 92
TA
a
01C 41,
-Knee-
kI
I
Figure 4.3.8.
Subject LLA, tests 0.85G/N/B2/1-3.
-
93
-
400 OSer
kcee.
Figure 4.3.9.
Subject TWA, tests 0.5G/N/B2/1-3.
-
94
-
100 ysec
iN
0
c00
e00
a~kl~)
0
0
Figure 4.3.10.
0
0
-_
Subject TWA, tests 0.85G/N/B2/1-3.
_ _
_ _ _ _
_ _ _
-
95
100mge~c
9L
4
4
TA
Rr
ee0
acce.
a
v0
0
00
4n0
Figure 4.3.11
Subject PPE, tests 0.5G/N/B2/1-3.
-
96
loo mseC
accel
ankle
Figure 4.3.12.
Subject PPE, tests 0.85G/N/B2/1-3.
-
97
Li-o±G
Figure 4.3.13.
TK>$cj
Subject PNE, tests O.5G/N/B1/1-3.
-
WAN
98
-
0
0
0a
FiA
acce-
Figure 4.3.14.
.Subject
PNE, tests 0.85G/N/B1/1-3.
-
99
-
rather
leads,
specifically to electrode
than
This distinction should be kept in mind because
themselves.
of the possibility that a given pair of electrodes
EMG signals from adjacent muscles.
detect
muscles
the
of
in significant crosstalk between pairs
may
also
This might result
which
electrodes
are adjacent, such as the GN and SL leads, specifically.
EMG
The first
response
following
release
generally
occurs with a latency of about 100 msec, where the latency is
defined relative to the
value
a
reaches
of
0.05g.
a
with
respond
of
latency
the
acceleration
The gastrocnemius (GN),
tibialis anterior (TA), and
(SL),
which
at
moment
biceps
about
femoris
120 msec.
soleus
(BF)
all
The earliest
rectus femoris (RF) reaction precedes these with a latency of
100 msec.
only
Within
these
individuals,
latencies
relatively constant, but they vary between individuals by
to
are
up
The RF response may precede the initial response of
10%.
the other muscles by 10 to 40 msec.
The amplitude of these initial reactions, which will be
to
referred
as the early response, is roughly a function of
the acceleration transient upon release.
the
afffect
amplitude,
and
Other
factors
these will be discussed later.
The amplitudes vary considerably for a given subject and
of
test
expectation
affect
the
conditions,
and
level
response.
suggesting
of
may
that
apprehension
factors
may
such
set
as
significantly
The very first drop for each subject
-
-
100
invariably elicited large responses in all
second
The
muscles.
of
occurred
had
recorded
and all subsequent drops rarely
drop
substantial
that
showed such elevated responses, indicating
habituation
the
following
immediately
the first
drop.
Many
of
a
manifested
soleus
the
EMG
early response, consistent with
double-peaked
the findings of Lacour
et
(analog-processed)
records
(1978)
al.
and
Vidal
et
al.
(1979) in the baboon.
The double peaked structure was highly
variable, though, and
could
was
a
test
test
of
independent
of
repetition
all
of
the
during
condition;
for a given subject, both
condition
types of structure could often be
time
in
seen
The presence or absence of such a double-peaked
recordings.
structure
be
not
seen.
The
10-msec
rise
the averaging filter should be fast enough to allow
of
resolution of EMG bursts separated by 25 to 35 msec, which is
the separation of the two components as reported by Vidal.
The period preceding
early
generally
response
the
landing
consists
of
but
a
following
variable
length
This
latter
quiescent period followed by another reaction.
reaction
characteristic features and timing for each of
has
the five muscles and depends upon the
actual
contact
similar
time.
It
will
acceleration
be referred
and
to
the
reaction
to
slower falls,
the
to as the late
Qualitatively, the reaction to the 0.85G falls
reaction.
the
is
except that the
-
-
101
sequence of events is compressed in time and
are
generally
Quantitative
higher.
will
differences
in the GN and SL tend to be the longest, with a late
periods
of
consisting
response
during
corellated
considering
that
a
The
contact.
preceding
GN
SL
and
entire
the
burst
large
drop;
of
are highl-y
responses
this
is
reasonable,
both muscles act as plantar flexors of the
(However, some
of the two muscles.)
and
SL
of
the
be
may
corellation
timing,
BF responses tend to coincide
due
proximity
with
the
but the details of the response are more
RF,
another
EMG burst may occur following a very short
quiescent period, and preceding the late response of
and
to
Immediately following the early response in the TA
variable.
and
just
activity
electrical cross-coupling because of the anatomical
GN
be
The quiescent
discussed in the statistical analysis section.
foot.
amplitudes
the
the
GN
Thus, during the late response period, GN, SL, and
SL.
BF are usually coactivated, with RF reactions suppressed;
RF
tends to be maximal during quiescent periods of GN,
activity
SL, and BF.
This agonist-antagonist pattern probably assures
adequate knee flexion.
Depending upon the subject,
coactivated with either GN and SL or with RF,
relatively
consistent
within
the
TA may be
but tends to be
reactions
of
a
given
subject.
Interpretation of the functional
and
BF
reactions
muscles acting
at
is
the
difficult
hip
significance of the RF
because
joint.
The
of the many other
knee
angle
data
102 -
-
during
as
limiter required
may be'a result of the knee angle
(This
fall.
the
a
extension
full
reaches
never
knee
the
contact, although
of the knee prior to
extension
gradual
indicates
actually
the
perhaps
device;
safety
normal
degree of knee flexion during a fall is less than that caused
by
device.)
the
plantarflexion
and
GN
activity
consistent
the
with
plantarflexion
the ankle angle, the gastrocnemius muscle also
by
indicated
is
While the SL
fall, as expected.
the
during
gradual
indicates
data
angle
Ankle
acts to flex the knee.
The fact
suggests
that
the
that
of
action
is
quadriceps
the
to
begins
actually
knee
extend
responsible.
Because RF activity in the early response preceeds all
other
recorded reactions, it is possible that hip flexion (which is
also an action of the rectus femoris) is assured by
prior
to
the
of the lower leg.
reactions
occur
to
flexion was observed
Significant hip
is
not
possible
conclusions about the interaction of
hip
flexion
reactions.
This
hip flexion might
flexion,
it
information
quantitative
in
result
especially
all
other
response suggests
musculature
prior
to
make
with
leg
is an interesting issue, though; increased
if
that
active
to
recorded
the
diminished
a
adequate to cushion the landing.
precedes
without
but
falls,
during
CNS
the
CNS
sending
ankle
need
plantarflexion
RF
The fact that
activity
may
during
adjust
commands
knee
for
the
is
activity
the
early
proximal
to the more distal
103 -
-
muscles.
after
Immediately
to
active
be
of
weight
contact due to the
continue
RF
the
The
body.
and
GN
SL
a short time to counteract the
for
The actions of
passive dorsiflexion of the ankle on contact.
the
active,
becomes
counteract the passive flexion of the knee on
to
presumably
contact,
RF, GN, and SL would all act to bring the person into an
position.
upright
contact.
following
TA
is
activity
An
TA response after contact
extended
often occurs when the person
variable
extremely
begins
to
on
backwards
lean
contact; both a large RF and TA response would be expected in
this case in order to dorsiflex the ankle,
and
bring
the
extend
the
knee
person forward into an upright position from
the backward leaning position.
To summarize the more salient features of EMG reactions
the
during
fall,
approximately
with
responses,
five muscles studied react to a drop
all
simultaneous
short
early
latency
followed by activation of the RF, coactivation of
the GN, SL, and BF, and a more variable reaction of the TA.
During the trials in which the subject was not dropped,
but
the visual field was moved unexpectedly (OG/U and OG/D),
few EMG reactions of
during
the
first
any
significance
occurrence.
In
were
the
twitch
even
the few trials in which
certain subjects showed small twitch reactions,
of
detected
the
latency
was about 300 msec and generally occurred in
-
all
five
of
the
characteristic
recorded
of
a
startle reactions
104 -
muscles.
This
pattern
is
generalized startle reaction, although
typically
have
shorter
latencies.
See
For subject EMA, the first trial of
figure 4.3.4 on page 88.
this test condition elicited
a
substantial
The
response.
second repetition shows no response, and the third repetition
shows only a
slight
twitch
in
reacted at all to these stimuli;
throughout the test.
RF.
Most
subjects
never
the EMG recordings were flat
Because of the lack
of
a
significant
response, the OG/U and OG/D cases are not shown in the graphs
of the data.
However,
the
simple
of
result
these
be remembered when interpreting the effects of visual
should
field motion in reactions during falls;
visual
stimulus
alone
is
not
it is clear that this
sufficient
to
not
elicit
the
However, this
characteristic EMG patterns seen during falls.
does
tests
preclude the possibility of subthreshold phenomena
at the segmental level.
4.5
Statistical analysis of the data
Because of the natural temporal separation of
responses
into
an
to
individually.
instant
at
EMG
early component and a late response, the
EMG was separated into two
order
the
determine
time
possible
segments
trends
for
for
analysis
each
The moment of release was defined
to
in
response
be
the
which the acceleration reached a value of 0.05g.
This usually resulted in
detection
of
release
within
one
-
sample
period
inspection.
of
the
105 -
release
time
as
judged
For the purpose of interpreting
EMG
by visual
latencies,
the 40-msec time constant of brake-deactivation should not be
overlooked.
contact,
Contact time was defined to be the moment of toe
determined from the ankle angle measurements as the
time at which the foot begins to rapidly dorsiflex.
Both the
moment of release and moment of contact, as well as all other
measures of the data, were
computer
program
and
computer values were
values
were
determined
threshold
spot
checked
the
the
detection
for
aid
schemes.
accuracy,
and
of
a
The
all
scanned to ensure that no gross errors occurred
due to artifact detection.
in
with
few
cases
Values could be entered
where
artifacts
interfered
manually
with
the
the
EMG
automated detection.
Four separate integrations were performed
waveforms
from
each
diagram illustrates
of
the
the
time
integrations were performed.
TrT
five
muscles.
intervals
over
The following
which
these
A typical GN reaction is shown.
Te"3Crnsec Tc
.4
on
>
Tct20m~sec
-
106 -
T. ,
release,
TV
were
clearly
+ 175 msec fell in
the
and
response,
respectively.
scanned and compared to
Te
to
to
computed
the
of
plots
third
The
These three
activation
early
the
pre-contact
late
the
the
Again,
.
muscle
attributable
due
activation
response,
late
quiescent period,
period.
net
of
activation
fall,
a
to
T
from
values provide relative measures
during
by
quiescent
this
performed
was
integration
separated
the
+ 175
and
early
which
msec, because in almost all cases in
responses
after
was chosen to be equal to Tr
T.
T.,
of
Next, the
.
just
time
to a
waveform was integrated from Tr
early response,
contact, T0
of
moment
the
to
moment
the
First, the filtered data was integrated from
original
values were
waveforms.
Integrations
over the specified time intervals were actually
approximated
by
summating
the
sampled
data
the
over
appropriate sample periods.
from
In addition to separating the early response
late
Tt
the
was
obtained
- 30 msec
should
EMG activity occurring at the moment of
response,
contact
reflect
to
Te
the
the
by
the
integrating
from
EMG
+ 20 msec.
These
integrated
values
to
adjust
effective
joint
ability
stiffness in preparation for contact and not reflex reactions
to contact, since 20 msec is less than the
shortest
latency
for spinal reflexes.
Before performing analysis of
variance
tests
on
the
-
of
measures
over
the
that
such
normalized
average
drops for each subject were equal to a
45
all
the values for each
activity,
EMG
integrated
subject were normalized
values
107
constant, K (arbitrarily chosen to equal 100):
(
the
is
X/
where
Xjcr/5
of
value
normalized
measure.
the
Differences in the characteristics of the electrode sites and
the relation between the electrodes, skin,
and
be
can
which
fat,
a
in
problem
insulating
because
subjects,
because
and
gain
the
which
factors
analysis
of
assumes variance homogeneity, the above normalization
seemed
of
Comparisons
reasonable.
of
of the fat overlying the muscle.
properties
Since these differences introduce
between
layers
subjects,
female
substantially reduces the recorded EMG amplitude
the
large
Subcutaneous
differences in recorded EMG amplitude.
of
for
responsible
be
may
muscle
underlying
fat
subcutaneous
EMG
different treatments was the main concern,
vary
variance
scheme
reactions between
and
not
overall
differences in activation levels between subjects.
Figures 4.4.1-20 on pages 108-127 present the normalized,
EMG
integrated
integration.
The
top
condition;
data for each of the four periods chosen for
Each page displays the
graph
the
on
each
first
page
three
data
for
one
muscle.
shows six bars for each drop
bars
represent
the
average
(cont. on p. 128)
- 108
-
RG/N
FDG/N
SDG/N
0.85G/DK
~
I
I
0.85G/EC
___________
_________________________________
n 85/U
________________________
)
0.85G/D
___________
U0.80G/
0. 5G/DK
0. 5G/EC
a-.
0.5G/U
0.5G/D
-
0. 5G/N
I
C3
___
(3j
Gastrocnemius,
Figure 4.4.1.
integrated EMG during entire fall.
-
t
109 -
*1 I
RG/N
FDG/N
______
I
SDG/N
0 .85G/DK
I
___
____
0.85G/EC
I
____
______
0.8 5G/U
I
0.85G/D
0.8 5G/N
I
_______
0.5G/DK
0.5G/EC
0.5G/U
0.SG/D
0.SG/N
Ii
_____I
Figure 4.4.2. . Soleus,
integrated EMG during entire fall.
_________
______
-
110
-
RG/N
FDG/N
SDG/N
-
0. 85G/DK-
0.85G/EC
..1______
_______
.1
I
0.85G/U
__
-i
___
0.85G/D
t
0.85G/N
I
I
0. 5G/DK
I
____
0.5G/EC
2
I
0.5G/U
.1
0.5G/D
0.5G/N
V
Q
Figure 4.4.3.
Tibialis anterior,
integrated EMG during entire fall.
_____
I
___
Ii
____
___
I
0
-
/
'*0______
ill
-
I
_____________
RG/N
b-i.'
FDG/N
I
SDG/N
_____________
0.85G/DK-
I____
-
_____
j
I_____
______________________
0.85G/EC:
0.85G/U
0.85G/D
*1
0.85G/N
0.5G/DK
-I
0.5G/EC
0.5G/U
0.5G/D
0.5G/N
in.
in
0
Rectus femoris,
Figure 4.4.4.
integrated EMG during entire fall.
el
Q
-
112 -
'-9.
-~___
RG/N
FDG/N
SDG/N
I
0.8 5G/DK
____________________
'1
0.8 5G/EC
I
.1
_______
0.85G/U
1
0.85G/D
1;
0.85G/N
0.5G/DK
I
,1_______________
__________
0. 5G/EC
0.5G/U
Ii
0.5G/D
0. 5G/N
0
0
a
C),
Figure 4.4.5.
Biceps femoris,
integrated EMG during entire fall.
Q
'I'
113
-
-
-i
___________________________
I
_____________
RG/N
FDG/N
SDG/N
1..
0.85G/DK
I
I
0.8 5G/EC
0.8 5G/U
I
_________
0.85G/D
I
0. 85G/N
__
0.5G/DK
0.5G/EC
~-Ii
0.5G/U
0.5G/D
0.5G/N
In
Q
4c
1:A
0
Gastrocnemius,
Figure 4.4.6.
integrated EMG during early response.
Q
-
1
___
114 -
_____
RG/N
FDG/N
SDG/N
-
0.85G/DK'
-K
I.
7
I
1P
0.85G/EC
______________________
0.85G/U
________
0.85G/D
<4 W
0. 8 5G/N
0. 5G/DK
i
________________________________________________________
J_______
0.5G/EC
0.5G/U
0.5G/D
'1I
0.5G/N
-
___
_________________________
0
v~I
Qo
0*
C
in
0
a
Figure 4.4.7. , Soleus,
integrated EMG during early response.
0
-
115 -
I
_____
RG/N
FDG/N
t
_____________________________________________________________________
SDG/N
0. 85G/DK-
I
________
j
0.85G/EC
________
0.85G/U
I.
I
0.85G/D
_______
I
0.85G/N
0.5G/DK
0.5G/EC
b-i__
0.5G/U
0.5G/D
I
I
K-
___
0.5G/N
V7
in
Tibialis anterior,
Figure 4.4.8.
integrated EMG during early response.
0
'A
6
-
i
I
116 -
I-
I.
0-~
RG/N
FDG/N
SDG/N
I
0.85G/DK
______________
I
0.85G/EC
j
_____
I
0.85G/U
________
0.85G/D
K~1
0.85G/N
0.5G/DK
r
_____
0.5G/EC
0.5G/U
71'
I
___
0.5G/D
0. 5G/N
COi
a
R
1
4~)110
Figure 4.4.9.
Rectus femoris,
integrated EMG during early response.
___
-
____
Ii
117 -
-i
I
__
RG/N
FDG/N
J
K-
SDG/N
0.85G/DK
I
_____
I
_____________
1~~j
0.85G/ECJ
4
'7
0.85G/U
0.85G/D
t&-ji
I
I
0.85G/N
_______
0.5G/DK
9
0.5G/EC
I
________
177
0.5G/U
0.5G/D
0.5G/N
'hfl
Figure 4.4.10.
Biceps femoris,
integrated EMG during early response.
10
0
- 118 -
*1
____________
._
I
RG/N
I
FDG/N
SDG/N
0.85G/DK'.
0.85G/EC-
I-.__
0.85G/U
0.85G/D
0.85G/N
'-4
0.5G/DK
.prj-i'
0. 5G/EC
0.5G/U
0.5G/D
0.5G/N
1g~
Figure
*1______
a
4.4.11.
Gastrocnemius,
integrated EMG during late response.
0
-
119 -
*I
I
___
___
~~.1
RG/N
_______________________
FDG/N
I
SDG/N
_____________________
________________________
___________________________
_______________________________
0.85G/DK
0 . 8 5G/EC
7
7____________________
0.85G/U
0~.~~!f
0.85G/D
0.85G/N
0. 5G/DK
I
___
0.5G/EC
0. 5G/U
0.5G/D
0. 5G/N
U9
Figure
4.4.12.
Soleus,
integrated EMG during late response.
0
1(1
c
- 120
-
I,
RG/N
FDG/N
SDG/N
0.85G/DK
0.85G/EC
0.85G/U
0.85G/D
0.85G/N
0.5G/DK
0.5G/EC
I
0.5G/U
______
0.5G/D
_________________________________________________________________________________________
0.5G/N
Q
a
Figure 4.4.13.
Tibialis anterior,
integrated EMG during late response.
0
-
121
-
RG/N
I________________
FDG/N
SDG/N
0.85G/DK
0.85G/EC
0.85G/U
0.85G/D
0.85G/N
I
-
I
I
0.5G/DK
I
__
1
_____
0.5G/EC
I
______
0.5G/U
71
0.5G/D
I
*1
'2
0. 5G/N
~i.
0
'Cl
0
~7!__
I
1
_____
Figure 4.4.14.
Rectus femoris,
integrated EMG during late response.
I _____
0
-
I
122 -
1
RG/N
-I
j
_____
FDG/N
SDG/N
0.8 5G/DK
K
-j_____
0.85G/EC
0.85G/U
7
0.8 5G/D
0.85G/N
0. 5G/DK
0.SG/EC
1
-'
0. 5G/U
i
0.5G/D
0.5G/N
Figure 4.4.15.
Biceps femoris,
integrated EMG during late response.
________________
-
123 -
I
_________________________
RG/N
FDG/N
4
I
__________
a____ ____
SDG/N
0.85G/DK
0.85G/EC
I
I
______
0.8 5G/U
I
____________________
I.____
I
___
___________________________________________
0.85G/D
I
I
___
Ii
________________________________
0.85G/N
______
0. 5G/DK
0.5G/EC
0.5G/U
0.5G/D
0.5G/N
Gastrocnemius,
Figure 4.4.16.
integrated EMG during contact.
~1_
124 -
-
I_
RG/N
-I_
__
FDG/N
t-.
SDG/N
lit
I
f__
0.85G/EC
I
_______________
J____
0.85G/DE
0. 85G/U
I
__
J
__
I
0.8 5G/D
I
I
i
________________
0.85G/N
I
I
______________________________________
0. 5G/DK
C7
*1
I
H-
0.5G/EC
I
-~_
0. 5G/U
0. 5G/D
0. 5G/N
0
__I__
Figure 4.4.17.
Soleus,
integrated EMG during contact.
__
0
-
125 -
_______
*
I.-.
RG/N
p
2
4
FDG/N
SDG/N
0.8 5G/DK
0.85G/EC
I
I
0.85G/U
Ii
liz
.
TRI
I
0.85G/D
4
0.8 5G/N
H
I
______________________________________________________________________
0.5G/DK
___
I
ii
1
_____________________________________
__
5.
IL-
0. 5G/EC
0.5G/U
___________________________________________________________
0.5G/D
0.5G/N
Tibialis anterior,
Figure 4.4.18.
integrated EMG during contact.
I
___
- 126 -
1~
1
RG/N
i
I
_________
__
__
FDG/N
SDG/N
0.85G/DK
0.85G/EC
it
I
L
0.85G/U
t
_
_____
K-I
7___________
0.85G/D
0.85G/N
v-El
1
I
0.5G/DK
7 _____________
O.5G/EC
a--i___
I..-,
I__
__
0.5G/U
0.5G/D
0.5G/N
Figure 4.4.19.
'1 I
Rectus femoris,
integrated EMG during contact.
I
I
- 127 -
4
'-I______
1
f
RG/N
71I.
FDG/N
SDG/N
I.-'
-I
'I
______________
0.85G/DK
0.8 5G/EC.
0.85G/U
0.85G/D
0.85G/N
II
-t
I I
KJ__
I I
____________
0.5G/DK
I ____________
__
0.5G/EC
0.5G/U
0.5G/D
0.5G/N
CI
0
Q
IL-I_
Cm.
Biceps femoris,
Figure 4.4.20.
integrated EMG during contact.
0
128 -
-
block 1,
three repetitions in
and
the
three
second
repetitions in block 2.
the
to
correspond
all five subjects for each of the
across
values
normalized
The lower graph
shows the average value over both blocks and all
of
drop for each block was eliminated from
the first
magnitudes,
In block 1, the first test was
the statistics.
and for block 2, the first test was RG/N/B2/1.
in figures 4.4.1-20, these two drops are
shown
0.85G/U/B1/1
In the graphs
in
the
top
are eliminated from the group for the purpose of
but
graph,
response
elevated
significantly
showed
subject
each
repetitions
Because the very first drop for
given test condition.
a
bars
averaging in the lower graph.
Early response latencies relative to release time
for
determined
each
threshold
using
muscle
techniques similar to the procedure for
contact
and
original
or
very
times.
were
recognition
recognizing
release
Computed values were compared with the
In cases where the early response was absent
plots.
small,
the
value for the latency was treated as a
missing datum.
In order to quantify synergistic relationships
pairs
for
T,
of
the
to T.
corellation coefficients were calculated
muscles,
following
,
was calculated
and
between
T,
three
to Te
time
.
(V
Tr
to Te
The corellation coefficient
as follows:
[E CN
periods:
(V'v~Jj]
,
129 -
-
where
Y.
and similarly for all other expectations.
the
sample
the
corellation
digitally
and
n,
n
are
periods defining the desired interval over which
was
represent
V
and
versions of the original sampled, analog
filtered
filtered EMG, V,
V
calculated.
V / (n)= h (n)*V(n),
V.
and
where
the
impulse response is 'shown below:
IID
-q
-3
0
.1 -1
1
Z
3
q
This digital filtering reduced high frequency oscillations
the
EMG
and
in
increased the computed correlation coefficient
when coactivation actually occurred.
It
the
must be noted that the probability
filtered
EMG
values
is
not
distribution
Instead, the
Gaussian.
distribution is one-sided and skewed toward values near
due
to
the
coefficient,
However,
if
nature
as
of
such,
of
zero
rectified
signals.
A corellation
a
Gaussian
distribution.
assumes
the computed values are interpreted as relative
measures of coactivation, the analysis of variance techniques
can be applied to these derived measures in order to test for
differences and trends among treatments
derived measure.
as
with
any
other
-
Another
coactivation
difficulty
coefficients
130 -
in
the
arises
interpretation
from
the
of
possibility of
significant cross-coupling between the GN electrodes and
SL
electrodes, specifically.
and
tendon and lie very
factor
soleus
close
the
Much of the corellation of the
signals from these muscles may result from the fact that
gastrocnemius
the
the
are both attached to the Achilles
together.
Again,
because
this
is probably constant throughout the testing, analysis
of variance
conditions.
should
yield
valid
comparisons
between
test
-
4.6
131 -
Reactions to different acceleration profiles
Before considering the effects of altered visual
in
differences
motion,
profiles
be
will
is
of acceleration
effects
main
as
that
evident
a
approximation, the faster the fall, the large r the
In figure 4.5.1
on
end
of
in
points
Because
consists
taken
are
figures
these
constant of the brake, the
effectively
of
first
the
100 msec
is
a
at
The data
directly
relati vely
from
time
slow
acceleration
of
of a curve with an ave rage slope which
is different for each acceleration profile.
100 msec
reaction.
velocity
of
the first 100 msec following release.
figures 4.4.6-10.
first
page 132, integrated EMG activity during
the early response is plotted as a function
the
in
shown
data
the
From
analysed.
it
figures 4.4.1-20,
the
field
convenient
measure
of
the
The velocity
at
suddenness
of
release.
Trends for all muscles are significant
the
entire
(P<0.001)
over
range of velocities at 100 msec, but comparisons
between only the slowest four velocities shown do not reach a
significant level (P>0.05) for GN, SL and TA.
three muscles seem to manifest a plateau
function
of
reached;
the
in
Instead, these
activity
a
suddenness of release, until a certain level is
average
response
during
0.85G falls
definitely much higher, especially for GN and SL.
is
It must be
emphasized that the data do not allow conclusions to be
about
as
made
absolute values of muscle force; only relative changes
-
132
-
a t4
.0
Ia60.
10 ob%
100 t
50
Sb6
0
0
1
1
.i
,21
VbQ
0. 56
Ve/loc+y (*/Sea)
st 100 m
at
0.
1q
OT
e
Integrated ENG during early response,
Figure 4.5.1.
normal visual field conditions.
- 133 -
among different drops can be compared.
velocities of 94 and 119 meters/sec, respectively
conduction
Yoshida,
and
(Wilson
average
have
tracts
reticulospinal
and
vestibulospinal
to
Since the fibres of the
over by 160 msec.
is
and
120 msec
100
about
at
begins
usually
The early response
velocity
similar, conduction
(Henneman, 1974e).
of
distance
based on a conduction
to
brainstem
cat),
would be only about 15 msec,
time
conduction
gastrocnemius
the
in
1969,
for
the
assuming
and
1.35 m
faster
axons
motor
that
Therefore, it seems reasonable
a
the
early response signal which is sent as a result of processing
and
in the brainstem
received
information
velocity
figure 4.5.1.
chosen
as
from
the
nuclei
might
vestibular
incorporate
receptors up to
The data indicate a strong relation
about 100 msec.
the
vestibular
and
between
the early response magnitude, as shown in
While the velocity at.the end of 100 msec
was
the coordinate for the abscissa because it seemed
reasonable, other
interpretations
are
possible.
certainly
Acceleration at the end of a certain time could also be used,
and a similar monotonic relationship would be observed.
important
The
point is that the response magnitude is a function
of the suddenness of the release.
The following table presents average latencies
early
response
recorded.
during
of
the
0.5G and 0.85G falls for all muscles
(Visual conditions have no significant
effect
on
-
134 -
therefore, the following values represent averages
latencies;
for all visual conditions.)
(latencies in msec)
0.85G
0.5G
GN
106t11
SL
110 t 12
99 2 7
104 t
9
TA
99
13
95 t10
RF
83
12
78 t 6
BF
99 t11
87 t 8
Using analysis of variance, the trends for both GN and SL are
statistically
not large.
profile
(P<0.05), but the differences are
significant
Deterministic
characteristics
due
differences
cannot
be
to
acceleration
ruled out as a possible
reason for the variation.
GN and SL early responses typically account for only 10
to
20%
of
the total integrated EMG during the entire fall.
On the other hand, TA and RF early responses represent 40
50% of the total for these muscles,
comparisons
relative
of
absolute
and BF,
muscle
25 to 35%.
force
are
to
While
not
possible from EMG data, these figure do indicate the relative
functional
muscle.
significance
of
the
response
for
each
GN and SL early response is, therefore, probably not
a major contributor to the total
muscles.
early
force
developed
in
these
- 135 -
Figure 4.5.2 is
differences
which
relevant
are
not
present,
because
of
the
small
but rather, because of the
striking similarity of the integrated activity values for all
acceleration
profiles.
It
must
be
values were obtained by integrating
remembered
the
filtered
from a fixed time to the moment of contact.
falls result in a shorter fall time,
that these
this
EMG
Since the faster
implies
that
EMG level must increase in order to achieve
average
total integrated value.
For example,
data
vs.
of
figures 4.3.1
4.3.2
the same
figures 4.3.5
4.3.6; these comparisons are between the O.5G/N
and
The normalized
EMG
period
for
the
late
response
are
vs.
0.85G/N
conditions in two different subjects.
levels
the
the. original
compare
and
data
average
shown
in
The integrated EMG during the late response period,
as
figure 4.5.3 on page 137.
shown
in
figures 4.4.11-15,
shows much less variation with
acceleration profile than during
the
early
response.
The
0.85G/N drops elicit somewhat greater late responses than the
other IN cases for all five muscles.
that
these
values
were
the
fastest
remembered
obtained by integrating the analog
filtered EMG from a fixed time
since
be
It must
to
the
moment
of
contact;
falls result in a shorter fall duration,
this implies that the average
EMG
level
must
increase
in
order to achieve the same total integrated value.
Integrated
EMG
during
contact,
as
shown
in
- 136 -
1604
-
.........
100 +
-------
go.
A
4
Oace .
fro9f
-. RG
Sb 4
-?. 26
O5<
1/SeC
'1
11
0.
Y5G(
Figure 4.5.2.
Integrated EMG during late response,
normal visual field conditions.
-
137 -
iSO4
100 1.
560
-
C&ACtC
Y..ekocifx"y
cc e I
-TA
I
pro40 e: j?
Figure
"e
"?/S c
24
?
SbF
bG
7
4.5.3.
Average integrated EMG (per unit time)
during early response,
normal visual field conditions.
- 138 -
figures 4.4.16-20,
indicates
consistent increases of muscle
activation in all muscles during contact as the speed of
fall
increases.
The
degree
reasonably be expected
contact;
in
for the /N
contact
to
of
muscle
correspond
contraction might
to
the
velocity
falls is plotted for all muscles as a function
the
of
Differences between pairs of drops for a
given muscle do not reach a statistically
for
at
figure 4.5.4, the integrated EMG during contact
velocity.
except
the
difference
significant
level,
the 0.5G/N and 0.85G/N
between
drops for all muscles (P<0.01).
Timing differences between 0.5G/N and 0.85G/N falls for
the
late response are evident in the original data from most
of the
subjects.
original
Compare
data:
the
following
figures
4.3.5 vs. 4.3.6,
of
4.3.9 vs. 4.3.10,
4.3.11 vs. 4.3.12, 4.3.13 vs. 4.3.14, and to a lesser
in
In
4.3.1 vs. 4.3.2.
figures
contact
words,
time,
the contact
80 msec
later
than
shifted
by
similar
a
relationship
which differs by about 80 msec.
time
for
the
the
0.85G
amount
0.5G
fall;
and
fall
purpose
contact,
of
comparison
since
reflex
to
the
It
is
ignore
any
activity
to
In other
occurs
about
the late response is
quiescent
following the early response is lengthened.
can be seen in the other pairs.
extent
4..3.5 and 4.3.6, the late
response demonstrates a relatively constant
the
the
period
Similar patterns
important
activity
following
obviously be related to the moment of contact.
for
the
following
contact
will
0
- 139 -
fr
7
too
t
-
-
-
- - - -
-
-
TA
- - - - - - --
z3.
S
'?, .1,
4
:
,cceG
0-5a
O.5L~e
Figure
4.5.4.
normal
R
Sb
ib,
Integrated EMG during contact,
visual field conditions.
-
140 -
The fact that almost all points for the
FDG
drops fall
the point representing TA during the FDG fall.
is
in
plotted
the
0.5G,
order
SDG,
and
changes
acceleration
waveforms.
during
For example,
occurs
transient
acceleration
the
during
end
until
late
Therefore, any
figure 4.1.
because
the
such
later
0.5G falls,
that
in
the
the
only
the initial release;
however, for the RG and SDG falls, the ramp
not
All
are
are
progressively
occur
to
2.2 m/sec
RG,
characteristics of the acceleration profiles
attempt
and
The only exception
of the falls which have a contact velocity of
does
RG,
in the range between the 0.5G and 0.85G drops
for all muscles, may suggest a trend.
this
SDG,
acceleration
in
into the fall, as can be seen in
internal
which
mechanisms
may
to predict final contact velocity would have to make
adjustments for this fact later in the reaction.
The results are consistent
acceleration
changes
with
during the fall
the
induce reactions later
in the fall, although the differences do not reach the
of
significance
for
during contact for the
contact
for
the same.
is
that
each muscle separately.
SDG falls
are
level
EMG reactions
greater
than
during
0.5G falls, even though the contact velocity is
Another result which can be deduced
the
subject
successfully
acceleration increases during mid-fall.
particular,
that
prediction
if
the
from this data
accounts
for
these
For the SDG drop, in
mid-fall acceleration increase were not
taken into account, the projected contact velocity
would
be
about
1.8 m/sec
instead
-
141
-
of
2.2 m/sec;
contact EMG levels
would then be expected to be less than for the 0.5G case, not
greater.
4_.
General habituation
0.85G/U/B1/1
In the test sequence of block 1, test
the
in
the
during
these
tests
are
the
second
and
significantly greater than reactions during
repetitions
third
the
test, sometimes by as much
same
However, the second and third tests of the sequence
as 200%.
for
of
block 2, test
of
In figures 4.4.1-20, almost all
reactions
measured
the
sequence
test
is the first test.
RG/N/B2/1
of
test;
first
is
both blocks elicited EMG responses which were within the
range of values for all
or
reduction
increase
throughout
the
indicating
that
subsequent
in
of
remainder
almost
general
the
No
tests.
significant
EMG amplitude was found
tests
in
the
session,
all of the habituation had occurred
before the second test.
The pattern of habituation of the integrated EMG during
the
late response and contact should be noted- in particular.
The GN and SL reactions even during the first
far
outside
session.
fall
are
not
the range of reactions for the remainder of the
The TA, RF, and BF do
show
considerably
elevated
late
response EMG during the first trial, and TA and RF also
show
considerably
pattern
is
elevated
strikingly
values
during
contact.
This
diferent from that of all subsequent
-
142 -
Possible reasons for
reactions.
this
pattern
peculiar
of
during the first trial will be discussed in the
coordination
next chapter.
4.8
Main visual effects
Drops during
visual
modified
indicate
motion
field
numerous significant effects on the response pattern,
but the
details of the changes are complex and do not allow a
simple
linear model of the influence of visual field velocity on EMG
reaction magnitude.
Visual effects are most clearly seen in
different
in integrated
differences
figures
GN,
0.5G falls.
4.4.1-5.)
While
SL,
the
and BF show the most dramatic
EMG during the 'entire
certain
significant levels in all of
comparing
fall.
(See
comparisons may not reach
these
muscles,
the
following
general trends may be seen: compared to the drops with normal
(IN) visual condition, reactions are diminished in the /D and
/EC
cases,
increased in the /U
affected by darkening the room (/DK).
response
in
the
/D
cases
and not significantly
cases,
The diminishing of the
is the most significant effect.
Significance limits as determined by analysis of variance are
shown
the
in
following
table
different cases for 0.5G falls
fall).
A
blank
entry
in
for
comparison between the
(integrated
the
table
comparison did not reach the 0.05 level.
EMG during
indicates
entire
that the
-
0.5G/N vs. /D
143 -
SL
<.01
<.001
/N
vs.
<.01
<.01
<.05
IN vs. /U
/EC
<.01
<.01
By examining the integrated EMG values
and
late
responses
BF
RF
TA
GN
for
the
for
the
different
early
0.5G falls
(figures 4.4.6-10 and 4.4.11-15), the early reaction
appears
responsible for contributing increased reactions for
largely
the /U cases.
The GN and
significantly
elevated
of
modulation
that
GN
any
values
fields
cause
/U
case
other
for
is
visual
/DK
the
case.
While
bidirectional
total response magnitude during the fall
the
to the normal case,
compared
the
for
SL
and
visual
non-normal
the various
above
fo'r
also elevated relative to the /N
are
condition
values
the
although
condition,
SL
the early response is
increased
by any kind of unusual visual field motion which differs from
the
tabulated
open
eyes
normal,
below
for
case.
cases
Significance
where
the
early
values
response
significantly greater than the response to a normal fall:
0.5G/N vs.
/U
/N vs.
/EC
/N
/DK
vs.
GN
SL
TA
<.05
<.01
<.01
<.05
<.05
<.05
<.05
are
is
-
144 -
During the late response period, the GN, SL, and BF all
show
similar
patterns
0.5G visual conditions.
the
in
the comparison between different
Since the late response accounts for
major portion of the total response during the fall, the
patterns seen in the integrated EMG for the entire
very
similar.
fall
are
In contrast to the patterns observed for the
early response, the abnormal visual field conditions tend
to
consistently diminish, rather than increase the late response
magnitude.' The only exception to this occurs in the RF
response;
if
any
differences are observed, the response is
increased relative to the normal case.
are
shown
below
late
for
significantly different
cases
from
in
Significance
values
which the late response is
the
late
response
during
a
0.5G/N fall:
GN
0.5G/N vs. /D
SL
<.001
TA
RF
<.001
<.001
IN vs. /U
IN vs. /EC
During
between
the
evident.
case;
SL
the
<.05
<.001
contact
various
<.001
period,
visual
few
major
conditions
EMG
during
contact
greater than during a normal fall (P<0.001).
greater,
differences
for 0.5G falls are
The most striking difference occurred for
integrated
but
BF
the
/DK
is significantly
GN EMG is
also
does not reach the 0.05 level of significance.
RF contact EMG demonstrates some modulation with
respect
to
-
field
visual
it
motion;
145 -
during the /U case and
increases
However,
decreases during the /D case.
the
only
case
/U
compared to IN is significant (P<0.05).
The effects of different visual field conditions on the
0.85G falls are not nearly as dramatic as for the 0.5G falls.
differences
to
compared
0.05G falls.
the
between
of
wealth
the
During
the
significant main effect is that
period, the only
response
the
for
during
observed
effects
early
conditions
field
visual
various
significant
of
In fact, the most notable result is the lack
/DK
and
0.85G/EC
falls, the GN and SL response magnitudes are reduced relative
Note
to the IN case (P<0.05).
reverse
of
that
is
this
exactly
the
the result obtained for 0.5G falls, in which the
/DK and /EC conditions elicit
an
increased
response
early
relative to the 0.5G/N condition.
as
During the late response period' of 0.85G falls,
the
early
response,
the main significant effect is that GN
and SL responses during /EC
for
GN during /EC
the
contact
differences.
and /DK
presents
a
of
number
EMG during
interesting
The SL values for the 0.85G/D case has slightly
elevated values relative to the IN
shows
(P<0.01
falls are reduced
and P<0.01 for SL during /DK).
period
in
significantly
case,
and
/U
case
lower (P<0.05) values than the IN case.
GN and SL typically show large bursts of EMG
to and during contact.
the
activity
prior
This result indicates that the upward
-
146 -
moving visual field may fool the person into
he
is
falling
faster
prematurely activate SL.
decline
than
suggests
a
pattern
compared to the SL response.
for
that
him
on
to
its
the contact EMG measure would then
indicate a reduced response.
clearly
causing
reality,
If the reaction is already
contact,
during
in
believing
contact
The TA response during
of reciprocal innervation when
The TA
decreased
is
response
the 0.85G/D case and increased for the /U case (P<0.05),
which is just the opposite of
the
observed
pattern
in
SL
contact reactions.
To summarize
varies
the
significantly
main
the
effects,
with
early
in
differences
response
the
initial
more
acceleration transient, whereas the late response shows
variation
with
respect
to visual field conditions.
Visual
field conditions significantly affect both the early and late
responses
for
0.85G falls.
variation
0.5G
falls but do not have as much effect on
The fact that
the
late
response
less
with respect to the initial acceleration transient
is not surprising; if the late response is a more
preparation
volitional
for landing, its parameters should be determined
by the expected velocity at contact with
the
shows
the
floor.
Since
corresponding range of contact velocities tested (2.2 to
2.6 m/sec) was considerably less than the corresponding range
of
initial
acceleration
transients,
consistent with this hypothesis.
the
results
are
-
4L2
147 -
Interaction effects
Except for the drastic reduction of the
the
following
Observation
adaptation effects were seen.
upon
their
lost
did so only during the first
usually
contact
subjects
the
The few subjects who
their landings quite well.
balance
of
that they were able to control
indicated
falls
the
during
magnitude
other habituation or
few
fall,
first
very
EMG
trial, and learned almost immediately the correct method
for
landing without losing balance.
This experiment was
allowing
designed
sequential
immediate
patterns
However, no consistent
with
effects
were
the
intention
to
of
analysed.
be
While
observed.
it
might be reasonable to suspect that the immediately preceding
only
test may influence the reaction, the data here indicate
that
any
possible
significance.
effects
do
not reach the 0.05 level of
and
acceleration
The main effects of
vision
were the predominant factors affecting all tests.
Certain longer term interactions were noticed, however.
For the purpose of analysing these effects of sequence order,
table 4.4 on page 84 also shows the average
for
each
set
example, since
sequence
of
of
three
the
two
the
tests
blocks
of
have
sequence position
in
block 2.
the
exactly
reverse
other, if the average sequence position of
test 0.5G/N is 17 in block 1, then it must be
17)
For
a given condition.
If
the
comparison
of
29
two
(46
minus
tests
is
148 -
-
the
an
effect exists, and may be due to the possibility
interaction
that
block 2,
and
block 1
between
different
significantly
may
reaction
vary
with
number
the
drops
of
previously experienced.
of
During the late response period, the comparison
and
for
SL
0.5G/U and also for 0.5G/N and /DK
and
0.5G/N
reveals what seems to be
and
increased
an
/DK
conditions
sensitivity
to
sequence.
This effect is surprising since it
effects
by
the
indicates
that
are not ignored as the testing proceeds, but
The
may actually be enhanced.
block
of
later in the testing
reactions
visual
/U
GN
test
condition
significance
values
interactions
for
comparisons are listed below (see figures 4. 4 .16-1
of
the
7
the
above
):
block x test condition interaction during late response
GN
0.5G/N vs. /U
<.05
<.001'
0.5G/N vs. /DK
<.05
<.01
Tibialis anterior
interaction
responses
decrease
also
show
a
few
Here, TA responses generally tend
with experience.
Refer to table 4.4 on page 84
to verify the actual sequences for block 1 and
significance
interactions
significant
effects, but most of the significant comparisons
are among the 0.85G drops.
to
SL
values
block 2;
the
for significant block by test condition
are shown below:
149 -
-
block x test condition interaction during late response
TA
0.85G/N vs. /U
<.05
<.05
IN vs. /DK
These are the same visual field
most
for
interactions
significant
which
conditions
SL during 0.5G
and
GN
the
show
Other interactions did not reach a significant level.
falls.
While the interaction effects do
linkage
between
GN
the
and
SL
not
indicate
direct
and the TA responses, the
overall trends indicate a diminution of the importance of
with
response
and SL.
and an increase in the importance of GN
time
antagonist
This would be reasonable considering the
relationship
TA
these
betweeen
Though statistically
muscles.
significant, the effects are not great, as
can
be
seen
in
figures 4.4.16-18.
4.10
Coactivation coefficients
Table 4.5
on
page 150
coactivation
the
presents
coefficients, as defined previously, for all pairs of muscles
recorded.
Several results are of interest
that
of
all
early response
coactivation
the
is
here.
The
fact
muscles respond simultaneously during the
reflected
in
the
high
of
values
coefficients for the early response.
the
GN and SL
show consistently high values, but much of this may be due to
electrical
\
crosstalk
as
mentioned
previously.
While
150
-
Coactivation coefficients
Table 4.5.
and 0.85G/N falls.
for 0.5G/N
muscle pair
0.5G
0.85G
late response
0.85G
0.5G
GN-SL
.80
.85
.75
.71
GN-TA
.58
.65
.23
.09
GN-RF
.43
.45
-.06
GN-BF
.44
.66
.31
.38
SL-TA
.57
.62
.27
.14
SL-RF
.34
.37
-. 12
-. 26
SL-BF
.45
.70
.39
.36
TA-RF
.45
.47
early response
TA-BF
.56
RF-BF
0a60
*
Table
0
0
*
0
0
0
4.6.
for
all
a
0
0
a
0
0
0
0
0
0
&
*
0
0
0
*
*
a
0
0
0
-. 18
-. 02
.51
.18
-. 05
.48
-. 09
-. 14
0
9
a
0
0
*
0
0
4
0
0
a
6
0
0
0
0
0
0
0
0
0
0
0
0
a
0
GN-TA and SL-TA coactivation coefficients
0.5G and 0.85G late responses:
visual conditions.
0.5G
IN
/D
/U
/EC
/DK
0.85G
IN
/D
GN-TA
.23
.23
.36
.17
.32
.09
SL-TA
.27
.30
.36
.19
.32
.14
/U
/EC
/DK
.28
.23
.06
.27
.26
.28
.10
.33
4
4
-
151 -
most of the early response values
falls
are
not
is
the
0.5G
and
due
0.85G
falls
(P<0.001).
This
the tendency to increase the degree of
to
knee flexion simultaneously with ankle plantar flexion
in
0.85G
significantly different, the GN-BF and SL-BF
values are increased for the
probably
for
early
in order to cushion the landing; knee flexion is
the fall
more crucial during faster falls.
the level of TA co-contraction with GN and SL
Likewise,
decreases
during
the
late
response period (P<0.01).
This
would allow a given level of GN and SL contraction to be more
effective in plantar flexing the ankle.
Table 4.6
on
page 150
presents
the
coactivation
coefficients during the late response for GN-TA and SL-TA for
all visual conditions during 0.5G
coefficients
were
the
with visual conditions.
all
of
the
and
0.85G
falls.
These
only ones which varied significantly
Note that
for
both
accelerations,
non-normal visual conditions except /EC tend to
induce increased co-contraction of TA with GN and SL.
-
152 -
153 -
-
CHAPTER 5
INTERPRETATIONS AND CONCLUSIONS
The
experimental
comparisons
of
the
design
main
of
effects
this
of
study
different
allows
rates
and
acceleration with results reported by other researchers,
also
allows
comparisons
reported data.
The
of
of
visual effects with previously
testing
procedure
differs
in
several
important respects from previous studies and the results have
made a critical reexamination of previous results necessary.
Melvill Jones and
Watt
(1971b)
performed
the
first
detailed examination of human responses to sudden, unexpected
falls.
and
(See
other
Chapter 1 for a more extensive
related
studies.)
The
Melvill
procedure differed from the procedure
their
subjects
held
unexpectedly released.
nearly
the
Melvill
The
instantaneously,
brake
time
whereas
for
constant
of
Jones
overhead
into
in
zero-g
was
occurred
in the present experiment,
deactivation
was
40 msec.
Jones and Watt reported an early response latency in
of
assumed
must be
that
which
the
fall.
was
This
relatively
invariant
that an acceleration
reached
(Greenwood
and
before
Hopkins,
the
with
is consistent with the 99 msec
latency in GN obtained in this study for 0.85G falls,
is
this
and Watt
in this study
bar
release
TA and GN of 75 msec, which
height
a
onto
review
threshold
early
1976a).
if
it
Ag of about 0.2g
response
is
triggered
For the ramp acceleration
154 -
-
108
case, GN and SL responded with average latencies of
Again,
respectively.
112 msec,
this
is consistent with a
initiated
is
Ag of 0.2g before an early response
threshold
and
by otolith-dependent brainstem mechanisms.
and
Greenwood
(about 0.2g).
2 m/sec2
to
subject
They used a
system
the
of
system
required
which
be suspended in a harness, and released him
following deactivation of an electromagnet.
a
that
reported
for eliciting a detectable early response is about
threshold
the
(1976a)
Hopkins
the acceleration transient at
counterweights,
release was relatively rapid,
used
Since they
similar
to
that
of
Melvill
Jones and Watt.
of
To further determine the sufficient.parameters
it
stimulus,
wjould
be
useful
to use carefully controlled
acceleration ramps of varying slopes.
the
hypothesized threshold
certain time limit.
a
long
period
response"
Ag of 0.'2g applies only within a
it
gradual
Ag occurs
over
would seem likely that no "early
Ramp accelerations
would ever manifest itself.
sufficiently
It is conceivable that
If the total cumulative
of time,
the
slope
to
determine
of
threshold
specifications were not tested in this study.
The magnitudes of the integrated
the
early
EMG
activity
during
response are also consistent with the findings of
Greenwood
and
magnitude
of
Hopkins
(1976a),
who
determined
that
the
the GN early response was a linear function of
155 -
-
the
acceleration
a
indicates
Figure 4.5.1
transient.
on
function of early response magnitude
monotonic
provide
not
statements about the linearity or non-li nearity
with
respect
to
response
acceleration
chapter .
transient, as discussed in the previous
data
of
of initial
measures
various
conclusive
make
to
information
enough
data
This
versus velocity at the end of the first 100 msec.
does
page 132
If
these
differ in any respect from the res ults of Greenwood and
Hopkins, it is
in
result
the
and
SL
responses
more that the TA, RF or BF res ponses for the higher
increase
communication)
is
than
would
indicated
be
personal
by
a
linear
magnitude and initial
response
early
between
relationship
(1980,
confirmed that the GN respo nse for one-g
has
greater
W att
0.85G.
acceleration drop condition,
falls
GN
that
acceleration transient.
Both Melvill Jones and Watt (1971b) and
Hopkins
Greenwood
and
(1976a) clearly determined that if subjects knew the
height of the fall
beforehand,
could
they
accurately
and
consistently adjust the timing of the late response such that
maximal
activity
in
Subjects
sensations
heights
of
contact
Greenwood
surprise
and
at
with blindfolds on.
that contact
times
would
contact
preceded
occur
actually
by
to
experiment
Hopkins'
being
40
dropped
from
140 msec.
reported
different
Their expectations seemed to be
sometime
occurring.
response and subjective sensations
between
the
range
of
The timing of the late
were
consistent
with
a
- 156 -
preparation
for
landing
which
was either too early or too
late.
The first three subjects
tested
at
in
the
addition,
for
normal
never reported a sensation
included
late
response
timing.
of
surprise
after
falls
which
acceleration changes after the initial release.
contingent
upon
incoming
that
late
response
If
timing
vestibular information during the
then the absence of subjective
suggests
were
visual field conditions, subjects
subjects are capable of adjusting the
fall,
study
heights of 28 and 43 cm, and the results indicate
almost perfect ability to adjust the
In
current
sensations
the more automatic mechanisms
of
surprise
involved in the
late response do not require the level of conscious awareness
is
which
triggered by a situation where expectation differs
significantly from reality.
Both the early
response
habituate significantly,
trial,
and
late
response
and
by
Greenwood and Hopkins.
Jones
some
similarities
to a general startle reaction.
analogy is also suggested by the fact that no early
occurs
when
the
person
initiates
and
The fact that the very
first drop elicits much larger responses suggests that
are
not
with the exception of the very first
confirming the results obtained by Melvill
Watt
did
a
fall
there
This
response
himself.
As
mentioned in the previous chapter, the RF, BF, and TA muscles
all
manifest
greatly
elevated
early responses compared
to
-
157 -
whereas
those during subsequent falls,
such
present
GN
a
which
component
not
do
fall
may
is qualitatively different.
An
first
musculature
thigh
increased emphasis on EMG activity in the
and
SL
This fact suggests that
dramatic differences.
the nature of the early response during the
include
and
tibialis anterior suggests that the action might be more
similar to a crouching or withdrawal reaction,
rather than an
enhanced preparation for an antigravity reaction.
section 1.3
to
(Refer
increased
by
would be characterized
GN
and
The latter
SL
activity.
discusses the action of these
which
muscles.)
repetition, and is
usually
after.
unexpected falls persists even
any
further
Rather,
misleading.
the
in
significance
origin of functional
reaction
the
during
repetitions,
seems
predominantly
of
to
respect
reaction
response
with
disappears
numerous
startle
early
reaction
triggered
reflexly
a
to
comparison
early
the
and
auditory
with
gradable
not
Because
intensity.
stimulus
and
habituates
quickly
stimuli
visual
to
reaction
However, a typical startle
be
would
to
be
a
vestibular
organization
of
preparation for landing (Greenwood and Hopkins, 1976b; Lacour
et al.,
1978).
(Watt,
personal
In addition, its presence in decerebrate cats
communication)
is
not characteristic
of a
startle reaction.
The
effects
of
different
visual
field
conditions
- 158
the
support
They certainly do not
not simple.
that
careful interpretation, since the results are
more
requires
-
idea
naive
a linear weighted sum of visual field velocity estimate
plus vestibular velocity estimate is internally generated and
used to control the landing response.
field
The only previous study of the effects of visual
on
conditions
rapid
motor
between
differences
A
(1979) on baboons.
completed by Vidal et al.
difficult,
procedures in this study will make comparison
to
addition
fact
in
were used rather than
baboons
that
of
They tested the baboons by dropping them at one g in
humans.
a
the
number
procedures and the
experimental
their
was
falls
during
reactions
position;
seated
being in a seated position obviates the
need for a late response in the leg
indeed,
found
they
that
any
muscles,
extensor
late
was
response
and,
rapidly
extinguished with repetition.
Vidal tested only three visual field conditions:
falls
in the dark, with a normal laboratory environment, and with a
lighted box surrounding the baboon's
with
comparisons
the
of
these
head.
different
One
difficulty
visual
conditions is that the normal case and the stabilized
case
involve
vision
completely different visual environments; more
variables than
changed.
field
Their
relative
visual
field
velocity
are
being
results indicate that the early response is
slightly reduced by a darkened visual
environment
and
even
-
more
by
the
-
159
stabilized field.
However, placement of a box
around the baboon's head presents a completely different kind
of
environment which may induce the baboon to preset
visual
the gain of the early response to
than
rather
testing
the
a
reduced
level.
Thus,
effect of reflex reactions to the
combination of vestibular input and visual field velocity, it
is
possible that the Vidal experiments tested the ability of
the baboon to preset the gain of the
reflex
depending
upon
conditions of which the animal was aware well before the fall
commenced.
to
be
The results do not allow any definite conclusions
made concerning the role of visual field motion cues,
specifically, in rapid motor responses.
The current study differs significantly from the
paradigm
cases
in
the visual conditions used.
described
different
in
relative
the
visual
previous
field
subject was completely unaware
During
The IN, /D, and /U
chapter
correspond
the
to
of which the
velocities,
before
Vidal
fall
commenced.
the setup period prior to the fall, care was taken to
prevent the subject from detecting cues which might allow him
to deduce which relative visual velocity was selected.
all differences between the /N, /D,
solely
since
to
visual
field
velocity
the
visual
field
pattern
and /U
Thus,
cases must be
due
effects during the fall,
was
exactly
the
same,
otherwise.
The results of this study are comparable
to
those
of
-
but
Vidal,
if
only
his test conditions are reinterpreted.
IN
the
different
significantly
current
falls,
findings
though,
quite
/EC
the
than
more
slightly
agree
In
this
Vidal's
for
all
of
is
not
respect,
the
results.
the means for the IN, /D,
similar
the
responses
but
condition,
(P>0.05).
with
the
reduces
condition
/DK
The
case.
conditions
/DK
and SL early responses compared to
GN
reduce
and
/EC
the
both
First, for 0.85G falls,
significantly
-
160
cases are all
and /U
muscles
indicates that visual field velocity does
For 0.85G
recorded.
not
This
significantly
It
seems more
likely that the stabilized vision case in Vidal's
experiment
alter
is
early
the
in
similar
response
effect
for
0.85G falls.
to the act of closing the eyes.
The
baboon possibly interprets the placing of a box over his head
as
signal
a
environment
voluntary
eye
same
the
/EC
perfectly
condition
of
of
however,
not
study, the results are
does
reveal
significant
visual field velocity on the early response,
primarily for the 0.5G falls.
are
current
the
consistent.
The current study,
effects
the
if
as being analogous to
"stabilized vision" case is interpreted
the
would
phenomenon
Therefore,
closure.
the
concerning
information
visual
is being eliminated;
during
occur
that
significantly
60% increase in
GN
and
relative to IN.
The /EC
different,
an
/D
The 0.5G/N and
80%
and /DK
increase
conditions
case elicits
but the /U
in
but
SL
a
activity
conditions actually increase
-
161 -
the magnitude of the early response in GN and
SL,
with
/DK
causing the greater increase of 50%.
Although
falls differs
the
paradigm
current
analogous.
quite
these
experiments.)
enhanced
CE)
The
normal
conditions
are
(N),
since
the
slightly
diminished,
did not
/U
degree.
induced
to
interesting
and
the IN, /D, /U
to
and
0.85G
to the Nashner and Berthoz
condition
enhanced,
rather
than
early response during 0.5G falls.
the
be
construed
as
it,
the
early
although
response.
If
more
normal
anything,
and Berthoz paradigm,
sway
to
in
see
a
they
not to a statistically significant
In this respect, the 0.85G falls are
Nashner
(S)
stimuli than 0.5G falls, the /D and /U conditions
enhance
diminished
stabilized
The results for the 0.5G
For 0.85G falls, which could
vestibular
conditions
field
analogous
are not directly comparable
results,
the
visual
of
Chapter 1 for a description of
(See
conditions tested here.
falls
unexpected
from the Nashner and Berthoz paradigm (1978)
induced antero-posterior sway, the
are
involving
one-g
these
comparable
since their subjects were
would
be
repeated
on
It
environment.
sway
to
experiments
subjects who have just entered a reduced-g environment,
as can be produced during parabolic aircraft
such
flight patterns.
162 -
-
Proosed model. fr
jJ1
early response generation
disparate
In order to make some sense of the seemingly
the
for
observed
trends
different
acceleration
possible
model
early
and
visual
proposed
is
field
conditions,
one
for
the
accounts
which
A diagram of the model is shown
differences.
under
magnitude
response
figure 5.1
in
on page 163.
sensor
acceleration
The
model
other investigators which indicate
of
results
incorporates
the
of
portion
that the saccules are probably responsible for generating the
for
signals
the
triggering
early
Watt (1980,
response.
personal communication) has reported that by placing subjects
in
a
reclining
position,
the
early response magnitude is
diminished considerably when unexpected falls
with
force.
the
bungee cords to replace the gravitational
of
use
simulated
are
is
The physical setup
the
to
comparable
vertical
situation,
but simply rotated 90 degrees into the horizontal
position.
Data
parabolic
in
obtained
also
flights
indicate similar findings.
the horizontal and zero-g conditions,
longer
feet,
with
biased
which
is
vestibulospinal
Other
a
conditions
zero-g
the
one-g gravitational
apparently
responsible
saccules
during
In both
are
no
force toward the
for
diminishing
reflexes.
reports
indicate
that
the
vestibulospinal
reflexes are also depressed when the neck is in an off-center
-
163 -
downward acceleration
relative angular
a(t) of body; onset at t=Tr
acceleration of
of visual field
triggers use of visual cues
cosG =head deviation
from vertical
Saccules
fTr<
+<
-
I-
) = reduction
K(
d,
d istance
from visual
field
pattern
factor due to
unbiasing of
otoliths in
off-vertical
position
j
T"< +< Tr+1OOse
Tr+ 100"Sc
+
a
+
estimate
of relative
linear accel.
of visual field
U
e
C
e
c
use
release triggers
+
c
release triggers use
/
previous
experience
Expected value
of expected value
Ks,
Ks
eyes closed
< Ks ,eyes open
1
-K
expected
value is
updated
following
each drop
0 < Ks<1
I A,
level of
0.2
anxiety
motivation,
anxiety,etc.
I
accel. (g)
1.0
(f ree fall)
magnitude of early
response
Model
Figure 5.1.
for determination of
early response
magnitude.
164 -
-
reduction
may
the head
such
that the predominant sensitivity vectors of the saccules
are
Therefore, it is possible that the
position.
be
no
First,
to two factors.
due
Secondly,
vertical.
in
Such
figure 5.1.
described
a
K( 6),
and
been
The downward
prior
is estimated by the otolith apparatus
response
early
as shown
has
nonlinearity
saccular
by Fernandez and Goldberg (1976a,b).
acceleration
the
the gain by the factor,
reduce
the
will
thus,
and,
saccules,
the
from
vertical will
the
to
relative
tilt
decrease the one-g bias of
additionally
0 is the angle of tilt
0 , where
the signal by cos
will reduce
in the direction of motion,
oriented
longer
simply tilting
to
quantity is labelled as a in
this
figure 5.1.
The visual portion of the model represents
for
increasing
acceleration
or
field
visual
relative
fall,
the estimate of downward
than
by
the subject is aware of his distance from
from objects in the environment.
shown
fast visual processing.
pattern spatial
during
frequency
and
the
higher
Prior to a
the
linear
in dotted lines as a reminder that it
whether such a transformation actually occurs
involving
if
pattern
The transformation from
relative angular field velocity to relative
is
a
otoliths.
the
strategy
acceleration
indicates
velocity
indicated
a
velocity
is not known
in the pathways
More experimentation with
pattern-to-subject
distance
unexpected falls would be required in order to verify
this possibility.
Implicit in the model
is
the
assumption
- 165 -
that both retinal slip velocity and information about current
estimate
an
eye velocity can be added to provide
total
of
visual field motion with respect to the head.
seem
that
indicate
to
of
is
it
constant,
'the
perceived
even
self-rotation
It was
inappropriate in these circumstances.
reported that "with the angular speed of
held
(1977)
al.
angular-to-linear
an
such
transformation occurs for perception
though
et
Wist
experiments by
Circularvection
a
visual
surround
of rotary self-motion
speed
increased linearly with increasing perceived distance of
results
These
surround."
are
the
of significance in analyzing
perceptual effects, but their relevance to rapid
reflex-type
reactions is uncertain.
Note that the values represented by a and b are
The inclusion of the f
values rather than functions of time.
functions
symbolizes
information
onset
is
the
single
assumption
that
only
sensory
from the moment of release to the early response
used
acceleration
to
calculate
estimates
for
inertial
the
and
visual
immediate
field
purpose
of
determining early response magnitude.
The next stage of
acceleration
estimate.
generated.
estimate
processing
from
the
involves
saccules
If a difference arises, an
error
comparing
to
the
signal,
the
visual
e,
is
In a normal fall with head erect, the estimate of
downward acceleration, a, should exactly cancel the
estimate
-
of
relative
visual
field
166 -
acceleration,
direction, providing no error signal.
specifically
in the upward
This argument does not
address the issue of otolith dynamics; however,
if the otoliths are modelled as a
second
system
of
with
b,
a
step
rise
time
order,
overdamped
only several msec, the
dynamics should not significantly affect the reasoning.
dynamics
of
The
central neuronal processing of otolith input is
not known in any detail.
The nonlinear function which provides the
signal,
c,
compensation
accounts for the result that a stabilized visual
field (/D condition) does not decrease the early response
much
(if
at all) as an upward moving field increases it, at
least for the 0.5G falls.
In the /U
condition,
visual
estimate, b, is twice the saccular estimate, a.
A
This generates an error signal equal to -a,
compensation signal.
nonlinear
which
causes
a
This compensation signal is then
added to the saccular estimate prior to
A
the
A
acceleration
large
as
further
processing.
function for generating the compensation signal
insures that any conflicts between the visual and
vestibular
estimates of acceleration are resolved such that the adjusted
estimate,
a + c,
may be increased
relative
to
but
not
since
the
reduce
the
In the model shown, the compensation function does
not
decreased
significantly.
factors cos 9
and
K ( 0)
This
will
is
always
reasonable
act
to
a,
estimate relative to the true value, a.
167 -
-
on
depend
this
but
K(&),
may
be
To test this
likely.
possibility, the same set of visual field conditions could be
tested
on
subjects
the Watt studies.
lying on their backs or in zero-g as in
Such
a
series
of
tests
may
indicate
whether the compensation process occurs prior to the stage at
which the K(E9)
case,
the
factor reduces the signal.
effective
gain
of
the
If
this
compensation
is
would
the
be
reduced.
The next stage of processing accounts for the
differences
the /N, /EC, and /DK condition.
between
that for 0.5G falls,
the early
response
in
and
GN
observed
Recall
SL
is
increased whereas for 0.85G falls, the /EC and /DK conditions
cause
a
decreased
condition.
early
response
compared
to
the
/N
These results suggest that prior to the fall, the
subject determines that the acceleration estimate will not be
sufficiently reliable and decides to provide his own estimate
based on past experience.
of
this
stored
value
Then, the release triggers the use
in
determining
the
early response
magnitude; following the fall, the expected value is
to
include
relative
the
most
weighting
figure 5.1.
is
recent
allowed
by
A continuous range of
the
KS
factor
in
Because the early responses for 0.5G/EC and /DK
are not exactly the same, the KS
to
event.
updated
factor is prohably adjusted
a value somewhere between 0 and 1,
rather than completely
ignoring either the current estimate or past values.
168 -
-
In the final stages
estimate
passes
are
may
be
certainly
a
of
some
combination
of
As mentioned
previously,
the
For rapid transients as in sudden steps of
the
threshold is probably around 0.2 m/seci.
The saturation level of one g
result
that
(9.8 m/sec
0.85G/U
falls
significantly increased early response.
adjusted
than
one g
) accounts for
did
not
For this
the
elicit
a
condition,
a + c,
would he much
Considering that downward
accelerations
acceleration
greater than one g.
greater
rate
tested in this study did not allow a threshold
acceleration,
the
of
da/dt, and the total acceleration change, La,
to be determined.
observed
acceleration
No dynamics are shown in the diagram,
within a certain time period.
accelerations
the
possible, considering that the threshold
function
acceleration,
processing,
through a function which includes threshold
and saturation levels.
but
of
(free
estimate,
fall) rarely occur in nature, it
makes sense for any neurophysiological mechanisms
to
regard
acceleration estimates greater than one g with suspicion.
Since
other
factors
such
as
level
of
attention,
anxiety, and motivation may also affect the response, a final
gain is included which depends on these factors.
While
the
multiplicative
taken
too
interpretation
should
not
literally, it seems more reasonable than an
since
both
the
variance
and
the
be
additive
magnitude
model,
of the early
response increase as a function of acceleration. Refer to the
upper graphs of
figures 4.4.6-10 on pages 113-117;
although
-
individual variance is
average
values
condition.
with
not
169 -
shown,
repetition
note
the
variation
number
for
a
The variation for 0.85G/N falls is
of
given drop
greater
than
for 0.5G/N falls, which can readily be seen.
In terms of the proposed model,
the
results
obtained
here are consistent with the general idea proposed by Nashner
and Berthoz (1978) that short latency postural reactions
be
may
diminished by visual cues if those visual cues indicate a
situation which is sufficiently different from
The
inputs.
concept
of
sensory
conflict
embodied in a model of circularvection
rotation
and
actual
the
has
during
expected
also been
visual
field
rotation of the subject (Zacharias and
Young, 1977).
5.2
Possible
physiological
explanations
for
visual-vestibular interactions during early response
Although
previous
physiological
experiments
unquestionably indicate visual-vestibular interactions at the
level of the vestibular nuclei, the Keller and Precht
1979a,b)
results are difficult to compare and reconcile with
the results of this study.
rolloff
(1978,
frequency
for
These
authors
found
that
the
visual angular velocity influence on
the vestibular nuclei neurones during horizontal rotation
about
0.25
lIz,
('r =0.64
sec).
Such
is
slow dynamics would
hardly be consistent with the rapid effects of
velocity on the early response during falls.
visual
field
- 170 -
One difficulty with making such
a
comparison
is
the
fundamental difference in the way visual field motion must be
processed during linear translations
and
rotation.
During
rotations of the body around any axis, a knowledge of retinal
slip velocity
and
eye
velocity
is
sufficient
relative rotation of the visual surround.
order
surround,
to
deduce
retinal
insufficient;
be known.
slip
The
Since
relative
linear
velocity
deduce
On the other hand,
the analogous problem during linear motion is
In
to
more
complex.
motion of the visual
and
eye
motion
are
eye-to-visual surround distance must also
the
person
is
presumably
aware
of
the
distance, it is not inconceivable that this information could
be continually supplied to any neural
involved
in
the
combination
of
mechanisms
which
are
visual field velocity and
vestibular cues.
In the rotational mode, the
influences
on
vestibular
slow
dynamics
of
visual
units is consistent with the idea
that the semicircular canals provide accurate information
higher
frequencies,
at
but need to be supplemented with visual
information at lower frequencies where their gain is reduced.
The
situation may be different in the linear mode; while the
otolith organs are certainly not limited by
response,
with
a
their
frequency
time constant of only several msec, there
may be a limitation in the
accuracy
of
the
transformation
from head linear acceleration to body linear acceleration.
- 171 -
Such
observed
on
teleological
results.
early
motor
motivation
effects.
to
steady
reactions
perform
neurons
responded to
during
not
falls
physiological
and
Thomsen
which
relative
state
does
alter
searches
(1979)
responded
linear
should
to
visual
provide
for
these
determined
that
otolith input also
field
motion.
units
Only
responses during 0.59 Hz sine wave stimulation
were obtained; although more data must be taken to
response
the
The existence of rapid visual influences
Daunton
vestibular
speculation
dynamics,
recorded
by
these
units
appeared
Keller* and
Precht
determine
analogous to the
during
horizontal
rotations.
For purposes of comparison, note that for 0.85G/U falls
in
this study, the relative angular visual field velocity is
120 deg/sec at the end of the first
Assuming
no
100 msec
of
the
fall.
eye movements occur in this time (which may not
be true) this would result in a maxium retinal slip velocity
of 120 deg/sec, which is well above the saturation range of 3
to
10 deg/sec
vestibular
reported
that
nuclei
Keller
units.
that perception
180 deg/sec.
and
However,
of
the
found
Brandt
their
et al.
(1973)
saturates
at
While it is unlikely that visual information at
absence
of
early
response
due
that
saturates.
velocity
at
100
to
any great difference between the
0.85G/N and 0.85G/U cases might be
visual
for
circularvection
100 msec has much effect on the
160 msec,
Precht
perception
to
the
possibility
This would be
172 -
-
unlikely if, for example, the early response parameters
already
determined
30 msec.
The best
on sensory input during the first
based
way
were
to
eliminate
potential
saturation
would be to increase the distance between the subject and the
visual field so that a given linear velocity would not result
in
such
high
angular
velocities
across the visual field.
Unfortunately, the laboratory room used here did not allow
a
more spatious display.
The search for a physiological basis
influences
transmitted
independently of the vestibulospinal system
level.
Most
of
triggered
the
responses
indicate
visual-vestibular
Daunton
interactions
vestibular nuclei which are consistent with known
control
reactions
evidence
strategies,
observed
for
visual
during
at
perception,
vection
and
stimuli.
at
and
the
oculomotor
slow postural
Convincing
rapid visual control of postural reactions via
the vestibular nuclei has not been forthcoming, though.
nonlinearities
the
recent physiological data,
including the results of Keller and Precht, and
Thomsen,
visual
It is possible that visual influences
and interact with vestibularly
segmental
fast
on the early response should not be restricted to
the vestibular nuclei.
are
for
inherent
The
in the results of the current study
indicate a much higher degree of processing than that seen in
the
relatively
linear visual-vestibular interactions in the
vestibular nuclei.
the
nonlinearities
In addition,
suggest
the nature and complexity of
that
vestihular
and
visual
173 -
-
influences are highly interdependent,
the
which
would
diminish
likelihood that visual influences are transmitted to the
segmental level independently of vestihular influences.
Tectospinal pathways
route
Maeda et al.
and
(1977) showed
optic
another
possible
could
that
stimulation
of
While
the
frog
tract induced IPSP's and EPSP's in spinal
motoneurones, it would be
effects
provide
rapid visual inflences on the spinal cord.
for
tectum
might
useful
to
determine
if
similar
be seen using visual field motion rather than
electrical stimulation.
The dynamics
of
the
responses
to
natural visual input would be extremely valuable information.
However, because of the difficulty of intracellular recording
at
the
spinal
level
with
unanesthetized
experiments may not be possible.
about
high
frequency
visual
animals,
such
Given the current knowledge
information
in
colliculus, it is possible that the superior
the superior
colliculus
may
be involved in rapid visual-spinal influences rather than the
vestibulospinal
pathways.
However,
because
of
the
differences between the frog and human visual systems,
proposal is only speculative.
The
tectospinal
many
such a
pathway
is
relatively more important in frogs than in hunans.
Maeda also showed
stimulation
summated
stimulation at the level
result
that
the
tectal
algebraically
of
the
and
with
motoneurones.
optic
tract
vestibular
Since
the
of Nashner and Berthoz and the results presented here
-
174 -
between
indicate much more complex interactions
such interactions must occur prior to
vestibular information,
the spinal level.
and
visual
This
suggests
that
if
the
tectospinal
tract is the pathway which conveys rapid visual influences on
posture, then it
reflects
the
visual cues.
signal,
must
transmit
complex
information
combination
of
both
which
already
vestibular and
In terms of the model, perhaps the compensation
c,
could be interpreted as the command siqnal which
is sent via the tectospinal pathway.
5.3
Additional experiments suggested by the
early
response
model
The proposed scheme for the
response
suggests
a
generation
of
the
number of additional experiments which
might verify the nature of particular elements in the
The
most
crucial
early
test
would
be
to
conduct
model.
experiments
analogous to those of Watt in zero g and horizontal positions
using
the
study.
visual
Because
significantly
field
the
reduced
motion
conditions
acceleration
tested in this
estimate,
a,
would
in these cases relative to the actual
acceleration, a, the saturation value of one g for the
estimate
would
result
in
greater
range
enhancement by an upward moving visual field.
these
cases
expected
to
an
be
upward moving field
significantly
increase
for
final
potential
Therefore,
in
(/U condition) might be
the
early
magnitude for fall accelerations greater than 0.5 g.
response
-
175 -
visual
angular-to-linear
should
successfully occurs, the results
cues
visual velocity
threshold
is
the
because
simply
reached,
not
independent
be
of
distances may result in a loss of
extreme
While
distance.
the
of
transformation
a
such
If
subject.
the
from
pattern
distance
the
estimator could be tested by altering
an
(or acceleration)
velocity
field
of
presence
As mentioned previously, the possible
velocity
visual
is unlikely that this would
it
become an issue in a normal environment where nearby
objects
usually present in the visual field.- This idea has been
are
invoked to explain the phenomenon of height vertigo, which is
by
(Brandt et al.,
1977).
the
All of
obviously
low-frequency
increased
characterized
of
experiments
sway
study
were
current
a normal one-q gravitational field,
in
performed
the
postural
and the hypothetical saturation level for the final
is
By
one g.
also
repeating
these
estimate
tests in hyper-g and
hypo-g environments, it might be possible to determine if and
how
quickly
saturation
the
value
adapts
to
the altered
'
gravitational field.
The effects of head tilt on the early response were not
tested
in
the
present
hazardous to test on
would
reveal
the
study
humans.
because
Similar
it
tests
was
judged too
with
monkeys
effect of various amounts of head tilt on
the early response, and might verify the
appropriateness
of
-
176 -
A
and K(O ) factors involved in the estimation of a.
the cos 9
The expected value mechanism which seems to operate
the
/DK
and /EC conditions was hypothesized on the basis of
only the current
necessary
to
set
eliminated
previous
an
example,
the
during the 0.5G/D
experiments.
that
For
from
course
of
verify
actually occurs.
time
in
testing
Further
tests
are
expected value determination
if
the
sequence,
0.85G
falls
were
the early response
and /EC conditions should
be
less.
The
of the decrease would depend upon the amount of
experience
which
is. required
as
input
to
the
expected value determination.
5.4
Synthesis of the late response
An analogous model of the late response would
not
provide
as
much
since the processing
which
are
not
insight
involves
known
probably
as the early response model,
higher
level
in great detail.
CNS
functions
Both the visual and
vestibular
senses
information
for the synthesis of the late response, as shown
obviously
contribute
significant
by the data.
Comparisons of the responses to the
stimuli
indicate
that
the
timing
of
five
the
consistently precedes the moment of contact.
subject
is
capable
of
successfully
acceleration
late
response
Apparently, the
monitoring
acceleration well into the fall, since the initial
the
transient
177 -
-
in
information
to
continues
acceleration
RG
and
FDG
SDG,
the
inadequate
provide
would
100 msec
first
the
during
where
conditions
change after the first 100 msec.
The advantage of using the SDG, FDG, RG, and 0.5G stimuli was
they
that
all
from even the original data
different
quite
0.5G
for
100 msec
initial
the
Since
SDG, 0.5G and RG
falls
which are evident in the
are
to
figure 4.1
SDG
data
original
probably
person
the
similar,
Timing differences
also
support
this
However, because the actual
range of contact velocities and contact times
(refer
and
total integrated EMG responses to the
Compare figures 4.3.1-3.
idea.
RG
the
for
the changes well into the fall.
monitors
the
of contact time would he much
estimate
different than in reality, especially
profiles.
monitored,
were
is
If only the
falls.
0.85G
and
timing
response
late
that
acceleration
of
projected
subject's
It is clear
have comparable fall durations.
on page 73),
it
determine at what time prior to contact
not
is
great
would be difficult to
information
arrives
too late to affect the response.
Visual influences on the late response tend to be quite
different
than
visual
influences
on
the
early response.
Since the early response latency is relatively constant,
the
response
is
manifested
simultaneously
in
and
numerous
muscles, the effect of visual field velocity as
a
of
However, the
the
amplitude can he simply characterized.
modulator
late response does not occur simultaneously in these muscles;
-
the
as
Visual influences may manifest themselves
response.
and
number
important aspects of the
are
timing
and
coordination
178 -
of parameters of the late response,
combination
between
coordination
and
timing,
amplitude,
including
any
muscles.
/U
For example, for 0.85G falls comparison of the
/D
reveals a timing difference which causes more
conditions
of the GN and SL activity to shift
for
period
the
/D
fact
The
case.
into
later
may
alone
information
the
no
that
contact
noticeable
suggests
habituation to these visual influences occurs
vestibular
and
be
not
that
considered
sufficiently reliable as to preclude the use of other sensory
modalities.
were accurate, the
information
vestibular
If
information.
visual
abnormal
the
subject might be expected to eventually ignore
Perhaps with longer periods of testing,
subjects might eventually learn to ignore it.
indication
No
of such habituation occurred in this study, though.
Visual effects on the late
exactly the opposite of the effect
The
/D
causes
condition
response.
Here, the
amplitude
rather
the tendency to
conditions
may
the
during
0.5G
cause a reduction of the response, which is
generally
falls
response
large
a
general
the
the
early
reduction
effect
than timing.
reduce
on
seems
to
response.
of
he
the late
on
the
One possible explanation for
amplitude
in
abnormal
visual
be that any detection of visual motion which
- 179 -
conflicts with the vestibular signal is recognized as a
motion
the
of
Such recognition may then
world.
external
cue,
reduce the influence of the vestibular
in
role of cognitive processes
the
increases
late
the
then
it
since
The possible
signal whose accuracy is doubted.
the
becomes
true
greatly
response
of strategies available, and any such
number
post hoc explanation must not be taken too seriously.
5.5
internal
of
Significance of results in terms of models
feedback during intended movement
The fact that no early response is seen during falls in
the
which
subject
initiates the release himself (Greenwood
and Hopkins, 1976b) suggests
reflect
would
early
would
may
knew
that
be dropped within several seconds, yet even this
knowledge is not sufficient to suppress the
This
response
In the present study, the subjects
actual input.
they
the
between expected sensory input and
discrepancy
the
that
be
consistent
with
constructs the expected sensory
early
reaction.
a mechanism which actually
input
at
the
time
it
is
expected; if the actual event occurs at the predicted moment,
the actual sensory input would agree with the expected input,
and
any
compensatory
predetermined based
unchanged.
correspond
In
to
on
motor
the
strategies
expected
input
which
would
had
been
proceed
the case of falls, this latter process would
the
self-initiated falls.
pre-contact
reaction
seen
during
-
Such a concept
of
180 -
comparison
between
predicted
and
actual sensory input has been invoked to explain a variety of
perceptual phenomena
The
most
occurring
frequently
eye movements do not
visual
world,
voluntary
movement.
cited example is that normal voluntary
result
yet
during
in
apparent
externally
imposed
movement
eye
of
the
movements
and
intended eye movements during induced oculomotor paralysis do
cause
apparent
movement of the visual field relative to the
subject.
Internal
phenomena
have been proposed by Sperry (1950), von Holst and
feedback
models
to
Mittelstaedt (1950), and Teuber (1966).
been
used
feedback,
to
describe
efference
Physiological
and
the
copy,
anatomical
process,
and
account
Various
for
terms
including
corollary
such
have
internal
discharge.
evidence for the existence of
internal feedback pathways at various levels of the CNS
been reported by Oscarsson (1967, 1970),
have
Thach (1970a,b), and
Evarts (1970a).
MacKay (1966) pointed out that perhaps the
comparison
be
between
interpreted
cancellation
or
as
process
expected and actual sensory input should
an
evaluation
subtraction
as
rather
than
a
simple
had been commonly assumed.
Cancellation would need to be extremely accurate in order
prevent
apparent
of
motion
during
to
voluntary eye movement and
other tasks such as tactile exploration.
These internal feedback models are
equally
applicable
-
181 -
interpreting both the perceptual consequences and further
to
between
motor commands arising from discrepancies
and
the present study it is a
In
inflow.
sensory
actual
predicted
motor command in the form of the early response which may
a
manifestation
of the hypothetical internal feedback.
The
cues
are
visual
nonlinear manner in which
vestibular
and
determine the early response magnitude suggests
to
combined
be
that MacKay's idea of evaluation rather than cancellation may
be
visual and vestibular cues are each
upon
In other words, both
more appropriate interpretation.
a
the
other,
It
algebraically summated.
of
kinds
than
rather
complex
evaluated
estimated
is perhaps
conditionally
independently and
noteworthy
that
the
evaluation of multiple sensory modes that
operate in perception may also apply to certain reflexes.
To further test
it
model,
would
be
of
the
certain aspects
example,
the
what
might
idea
of
internal
the
to
interesting
sensory
happen
if
cues
the
make
motion
only
For
unpredictable.
subject initiated the
release himself, but would not know the nature of the
field
feedback
visual
and/or acceleration during the fall? How would
the early response be affected during unexpected falls if the
subject
were
informed
of the nature
of
the
visual field
motion and/or acceleration stimuli prior to the fall?
-
182 -
-
183 -
-
184 -
CHAPTER 6
SUMMARY
this
The results of
of
existence
during
influences
reactions
during
This is consistent with studies of
visual
visual
unexpected falls.
conditions of
conclusions
postural sway in humans
It also
about
rapid
did
allow
not
definite
visual effects on motor responses,
since the results could be explained by the possibility
the
baboon
subjects
of
However, the experimental
study
latter
this
(Nashner
expectations
the
confirms
(1979) in baboons.
al.
et
EMG
on
influences
induced
and Berthoz, 1978).
Vidal
the
demonstrate
clearly
thesis
could
that
their responses based on
alter
visual cues detected prior to release.
During
relative
gastrocnemius
rectus femoris
(TA),
profile
acceleration
both
(GN),
(RF),
and
EMG from surface
visual field motion were altered.
leads over the
anterior
falls,
the
soleus
(SL),
tihialis
and biceps femoris
(BF)
was recorded and analyzed.
Most of
the
statistically
significant
observed in the lower leg muscle reactions
GN and SL leads).
The major findings of
effects
the
(specifically,
the
were
present
study
are outlined below:
1)
EMG patterns in GN and SL leads are consistent
patterns
with
the
observed by Melvill Jones and Watt and by Greenwood
-
and Hopkins.
185 -
The early response was found to occur
degree in all of the muscles recorded.
to
some
The late response was
more variable and was generally timed to
correspond
to
the
moment of contact.
2)
The early response magnitude an all five muscle leads
an
increasing
function
of
the
suddenness
measured by the slope of the acceleration
is
of release (as
during
the
first
100 msec).
3)
Differences
magnitude
of
of
visual
the
early
field
motion
response,
but
did
not
affect
in
a
manner
consistent with the results of Vidal et al. .(1978).
maintained
are
the
It
is
that the stimulus conditions in the present study
methodologically
interpretation
of
consistent,
the
and
therefore
the
rapid visual influences on the early
response are more straightforward.
The
major
effect
seen
here was that for 0.5G falls, the early response magnitude is
generally
increased
increased
relative
case).
This
Downward
condition.
to
effect
visual
was
did
magnitude
upward
the
field
relative to body)
response
when
visual
field
motion
is
subject (compared to the normal
not
observed
motion
not
for
0.85G
falls.
(visual
field
stabilized
significantly
alter
the
early
visual
field
relative
to
the
normal
Late response magnitudes were diminished
in
the
affected
the
case of 0.5G falls, however.
4)
Visual
amplitude
field
of
motion
also
the late response.
significantly
Again, the effect was much
-
186 -
greater for 0.5G falls than for 0.85G falls.
5)
During falls in which the eyes were closed
was
the
early
response
0.5G
falls
and
darkened,
increased for
compared
The late response was not
0.85G
for
decreased
suggests
mechanism which uses a neurally stored,
by
affected
significantly
result
This
conditions.
GN and SL leads was
in
falls
eyes open visual field condition.
normal,
the
to
room
the
or
the
operation
expected
these
of
pattern
a
in
cases where visual information is missing.
A model for the synthesis of
and vestibular cues.
to
shown
nature,
Because the
primarily
be
a
such
of visual
aid
in
has
response
otolith-dependent
may
model
early
was
response
been
and reflexive in
the
understanding
corellates of such interactions, presumably at
physiological
the brainstem level.
self-initiated
significant
early
account for the complex interactions
to
developed
the
falls
The absence of the
response
in
further suggests that it may provide a
regarding
clue
early
the
organization
of
movement
control.
The rapid dynamics which are necessary to
account
for
the effect of vision on the early response are not compatible
with the relatively
slow
nuclei
response to visual input, in cases where
neurons
in
dynamics
the animal is held stationary.
dynamics
of
(It is
observed
possible
vestibular
that
these
are somewhat faster when visual-vestihular conflict
- 187 -
is reduced.) In addition, the complexity of the
indicate
that
a
simple algebraic summation of visual field
acceleration and
cannot
gravireceptive
account
possibility
for
that
independently
estimates
the results either.
visual
conveyed
and
of
acceleration
This precludes the
vestibular
information
visual
and
vestibular
On
information
in the tectum as well as in the vestibular nuclei,
it is possible that units in the tectospinal tract
a
are
to the spinal motoneurone level.
the other hand, because
converges
interactions
pathway
for
mediating
represent
rapid visual influences on spinal
cord motoneurones.
By altering the acceleration profile,
to
determine
that
the
it
The
late
possible
early response is a function of the
magnitude of the acceleration transient prior
response.
was
to
the
early
response is more complicated, and both
its amplitude and timing can be adjusted prior to contact
order
to
anticipate
requirements
during
contact.
Acceleration appears to be monitored well into the
incorporated
into
the
late
response,
possible to determine beyond what
time
in
fall
and
although it was not
further
information
was not incorporated into the pre-contact response.
Differences
significantly
in
affected
visual
the
late
field
conditions
response.
Many
also
of
the
effects observed during the late response were different from
those
seen
during
the early response, and suggest that the
-
late reaction is compensatory
greater
degree
of
are
in
nature,
voluntary control.
interpretations of the late
perspective
188 -
much
response
more difficult.
must be interpreted phenomenologically.
and
is
under
a
Because of this, any
from
a
physiological
Instead, the results
-
189 -
-
190 -
APPENDIX A
REVIEW OF POSTURAL TESTS OF VESTIBULAR FUNCTION
Note:
Specific
are
section
particular
with
references
the
at
listed
a
to
relevance
special
of that
beginning
See table B.1 on page 196 for abbreviations used in
section.
this appendix.
A.1
-o behavioral tests
Advantages and disadvantages
Vestibular
dysfunction
long
has
tested
been
by
relatively simple techniques which either require the subject
walking
tasks,
to
response
an
involve
or
externally
of
number
of
observation
applied
and
standing
the
disturbance
subject's
the
to
While such tasks are of value because of
senses.
vestibular
a
during
to maintain his balance
their simplicity, evaluation of the results is complicated by
the
involvement
of other sensory
possibility of learning, and
systems
involved
the
systems in such tasks,
complexity
of
the
motor
Visual, tactile,
in making the responses.
proprioceptive, and even auditory cues are important, and
if
have disorders of any of these senses, this must be
subjects
known.
the
Humans
respond
to
an
are
clever
adaptable,
designed
experimentally
totally
creatures
obviate
the
who
disturbance
need
to
use
may
with
the
strategies
that
particular
sensory or motor capability that the experimenter
is attempting to test.
All of
these
difficulties
must
be
-
191 -
considered when devising a test and interpreting the results.
the
that
is
problem
Another
to
response
subject's
a
particular disturbance may not be representative of "natural"
strategies employed during normal activities, and, therefore,
only
should
used
be
with caution to derive a more general
model.
A.2
Common stimuli used to proyide gross disturbances to the
vestibular system
Graybiel (1974)
Kornhuber (1974a)
galvanic,
Rotatory,
to
most
common
rotating
the
commonly
are
stimuli
The
disturbances of the vestibular system.
induce
used
and caloric
procedure
for
the
rotation
requires
method
subject about the vertical axis, such that the
horizontal
canals are approximately aligned with the plane of
rotation.
Since the canals act as integrating accelerometers
with a finite time constant, the response decays after
30
seconds.
rotation
Nystagmus
will
and
If the rotation is suddenly stopped, a sense of
be
induced
postural
in
for
example,
direction.
opposite
the
occur,
asymmetries
recorded as a function of time.
include,
about
Such
which
postural
can
be
asymmetries
the tendency to rotate the torso and
arms in the direction of perceived rotation.
Galvanic stimuli
involve
current
electrode placed at or near the ear,
injection
into
an
which results in a gross
-
discharge
of
the
192 -
vestibular
receptors.
non-specific nature of the stimuli,
Because
Caloric stimuli allow the advantage
ear
at
a
time.
the
explanation.
of stimulating only
The stimulus strength can be adjusted
simply by controlling the temperature of the
irrigate
ears.
Water
which
is
temperature causes convection curre-nts
water
colder
in
the
three canals may be stimulated by
this
used
than
method,
to
body
canal
which are opposite to those produced by hot water.
fluid
Since all
the
actual
of induced rotation is due to the combined effects
direction
of each canal, and cannot be readily deduced.
however,
the
the results are difficult
to interpret beyond a simple phenomenological
one
of
Brief stimuli,
tend to affect primarily the horizontal canals.
discussed later, many of the induced postural
nystagmus
have
predominant
components
in
movements
As
and
the
horizontal
by.
vestibular
direction.
A..
Individual
limb
stimulation
Peitersen (1974)
Fregly (1974)
Rotatory,
galvanic,
and caloric stimuli all
result
in
detectable postural disturbances, but predictable, repeatable
results are often difficult to
techniques.
Often,
the
obtain
direction
with
of
tendencies depend on many other factors.
many
observed
Some
of
of
these
postural
the
more
193 -
-
repeatable observations will be reviewed here.
As a general rule,
direction
of
the
laterotorsion (head-turning)
slow
phase
nystagmus whenever it occurs,
galvanic,
or
caloric
of
nystagmus
whether
stimuli.
caused
If
the
in
the
accompanies
by
rotatory,
arms are extended
forward, they also tend to drift in the direction of the slow
This is a less pronounced
phase.
occurs
among
lowering
of
subjects.
the
Cold
effect,
caloric
ipsilateral
arm
and some variation
stimuli
and
contralateral arm as well as the lateral
result
raising
drifting
in
of
the
tendency.
Hot caloric stimuli cause the reverse phenomenon.
Trunk reactions must be tested
seated
to
eliminate
while
the
subject
the effects of the leg response.
is
Cold
caloric causes the trunk to lean ipsilateral to the stimulus,
whereas hot water causes contralateral leaning.
Consistent patterns in the lower limbs
most
difficult
to
parameter.
the
During
speed
slow
of
walking
forward
blindfolded subject drifts
in
component
Fast forward walking
of
nystagmus.
the
direction
zig-zag pattern with no particular drift
given
direction.
been
the
During walking tests following
obtain.
caloric or rotatory stimuli,
important
have
opposite to the slow component.
walking,
of
the
a
slow
results in a
tendency
During slow backward walking,
an
is
toward
a
the drift is
-
194 -
During in-place stepping tasks, subjects usually rotate
in
slowly
the
must be taken to eliminate a
large
number
must
headphones, to
auditory
eliminate
legs,
the
the
with
cues.
The
To test
cause
legs
to
predominantly
in
trunk
the
effect
the
If the arms are
arms should be at rest.
the arm drift may
extended,
provided
directional
his own thermal footprint on the floor.
on
must
since a barefoot subject can detect
must wear shoes,
subject
masking
noise
or
soundproof,
made
that
so
room
the
Then,
Closing the eyelids is not sufficient.
be
cues,
patterns in the room are not detected.
illumination
general
other
blindfolded
securely
be
of
may seem initially.
which makes the test more tedious than it
First, the subject
However, care
of the slow phase.
direction
and
follow.
A.4
Normal and induced .w ay during stand
Nashner (1970, 71, 73, 76, 77, 78)
Nashner and Wolfson (1974)
While standing, normal subjects sway
the
anteroposterior plane,
be reduced
by
with some lateral sway,
isometrically
tensing
the
leg
which can
muscles
to
A cold caloric causes the subject
provide greater stiffness.
to sway toward the irrigated ear whereas a hot caloric causes
the
opposite.
during
standing
patterns.
Turning
can
the
also
head from the neutral position
alter
or
abolish
these
sway
-
195 -
Nashner and Wolfson (1974) used galvanic stimulation to
initiate
disturbance
a
to
the vestibular system while the
The
blindfolded subject stood on a servocontrolled platform.
platform allowed suppression of ankle proprioceptive feedback
by
the
ankle
angle
were
forced
to
keeping
strategies
Transient postural sway
resulted
facing
only
when
forward.
in
constant.
rely
the
Thus,
corrective
on vestibular mechanisms.
antero-posterior
direction
the head was turned and not when it was
This
was
construed
as
evidence
for
'a
head-body transformation of the descending motor commands.
EMG response in the gastrocnemius followed the onset of
the
galvanic
stimulus
with
a latency of 100 msec.
produced by the muscle followed with a latency
and
lasted for 350 msec.
was employed,
some
probably
to
due
force
feedback.
began
to
1300
msec.
It
was
vestibular information summate,
suppressing
the
vestibular
A_5. The tipping test
Peitersen (1974)
at
begin
concluded
instead
of
information,
phase (before 600 msec) was unaffected
absence of ankle feedback.
msec
600
msec,
When ankle feedback was
eliminated, the corrective phase did not
1100
200
In the case where no servo-control
corrective
ankle
of
Torque
by
until
that
after
ankle
ankle
and
feedback
since the initial
the
presence
or
- 196 -
The tipping test requires the subject to
cannot
People with destroyed labyrinths
platform.
maintain
In
will fall without the use of vision.
and
balance
his
head down on a
on hands and knees while he is tilted
balance
their
maintain
basal ganglia disorders, the tilt reaction is often
lacking.
resistance,
a
passive phase during which the subject starts to fall, and
a
A
originate
from
and the last
"stiction",
it
to
due
which
muscle
corrective
was
fall and maintains
hypothesized
to
As the
rapidly
arising
from
maintained
otolith
stiffness,
muscle
increased
characteristic
is
length
phase
to
more likely that the initial
seems
changes of muscle length.
gravity,
phase
attributed
was
phase
simply
is
the
labyrinthine feedback of a transient nature,
However,
feedback.
arrests
first
The
posture.
stable
which
state
steady
final
phase
response consists of an initial
tilt
normal
of
stiction
small
is
increases
vestibular
tipping test is not used much clinically,
or
amplitude
overcome
by
until the final
feedback.
The
because other tests
are presumably simpler and more sensitive.
A.6
Effects DL vestibular disturbances on segmental reflexes
Kots (1976)
Reschke et al.
Segmental
stimulation,
(1976)
reflexes
which
is
are
also
affected
by
consistent with electrophysiological
results demonstrating potentiation of the segmental
following
VST
caloric
activation.
Cold
caloric
reflexes
increases
the
- 197
ipsilateral
patellar reflex,
induced by striking the
tendon.
Hot caloric increases the contralateral patellar reflex.
The
fact that the reflex response to stimulation of the posterior
tibialis
nerve is not increased
Y'
indicates that the
is probably activated by the caloric
stimuli,
system
increasing
the
magnitude of the gastrocnemius reflex.
Prolonged
significant
exposure
effect
on
and
are
weightlessness
the
Soleus-spinal H-reflexes
zero-g
to
strength
of
are potentiated
inhibited
during
also
spinal
during
has
a
reflexes.
periods
hyper-g (Reschke
of
et al.,
1976).
A1
Tests of balance during walking
Fregly (1974)
Fregly
subjects
to
has
a
developed
maintain
variable difficulty.
their
series
of
tests
requiring
balance under circumstances of
The tasks include standing with feet in
tandem heel-to-toe position (modified
Romberg test),
standing
on one leg, walking heel-to-toe on a beam, and walking
line
on
the
a
Some of the tasks were done both with
floor.
In
eyes open and eyes closed.
objectivity
on
to
addition
simplicity
and
of scoring, the tests were required to provide a
wide range of difficulty and a high sensitivity to individual
differences.
By
effect of vision
proprioceptive,
repeating
can
be
tactile
the tests with eyes closed,
estimated,
and
auditory
but
the
cues
presence
the
of
in addition to
-
vestibular sensation still
to occur.
198
permits many complex
As a diagnostic technique,
analysis of a complex system
research
-
tool,
may
be
such a multidimensional
of
value,
but
as
a
it leaves too many variables unrestrained to
allow an understanding of basic mechanisms.
the walking tasks,
interactions
especially,
Performance
in
are susceptible to improvement
by learning and depend significantly on motivation.
-
199 -
- 200
-
APPENDIX B
THE FUNCTION OF THE VESTIBULAR SYSTEM
AND RELATED BRAINSTEM STRUCTURES
IN THE CONTROL OF POSTURE
Note:
Specific
references
section
particular
are
with
listed
special
the
at
relevance
beginning
to
a
of that
section.
B.1
General structure
and
innervation
Qf
the
vestibular
nuclei
Brodal (1974)
Figures B.1-5
anatomical
it
may be
contain
connectivity
useful
to
schematic
representations
of
with the vestibular nuclei to which
refer
throughout
the
text
in
this
appendix.
The vestibular nuclei include
on
nuclei
each
side
fourth ventricle.
included.
superior (SVN),
the brainstem on the floor of the
vestibular
nuclei
nuclei,
the
nucleus
supravestibularis.
are
the
control,
foremost
also
descending
and
interstitial
vestibular nerve, cell groups f, 1, x,
of
are
includes the
here
lateral (LVN), medial (MVN),
vestibular
(DVN)
distinct
A number of smaller cell groups
term
The
of
four relatively
nucleus of the
y,
and
z,
and
the
Of these, the LVN, MVN, and DVN
interest
for
considering
postural
since these areas contribute significant descending
-
Table 13.1.
201 -
Abbreviations used in text
bvFRT
bilateral ventral flexor reflex tract
DSCT
dorsal spinocerebellar tract
DVN
descending vestibular nucleus
FRA
flexor reflex afferents
iFT
ipsilateral forelimb tract
LVN
lateral vestibular nucleus
1VST
lateral vestibulospinal tract
MLF
medial longitudinal fasciculus
MVN
medial vestibular nucleus
mVST
medial vestibulospinal tract
RST
reticulospinal tract
SVN
superior vestibular nucleus
VSCT
ventral spinocerebellar tract
VST
vestibulospinal tract
202
-
-
vestibulocerebellum:
flocculus
nodulus
caudal uvula
mossy fibres
ventral paraflocculus
ventral tip of dentate
nucleus
dorsolateral reticular
PRIMARY
VESTIBULAR
arformation
-----------
AFFERENTS
2.
utricle,
sccule
canals-
Figure
B.1.
SVN
canals
ventrorostral
...
-,.
LVNJ
rostrolateral MVN
utricle---
. caudolateral MVN
saccule -
\ rostral DVN
utricle -
.y ventromedial DVN
Primary vestibular afferents to cerebellum.
-
203
-
vestibulocerebellum:
flocculus
nodulus
caudal uvula
ventral paraflocculus
e/
fastigial
nucleus
I
ipsilateral mossy fibres
collaterals
Vestibular nuclei areas
including:
ventrolaterocaudal DVN
group f
group x
ventrocaudal MVN
Figure
B.2.
Secondary vestibular afferents to cerebellum.
-
flocculus
204
generally
ipsilateral
nodulus
Purkinje
axons
-
areas of vestibular
nuclei also receiving
) primary vestibular
afferents
peripheral SVN
caudal, medial MVN
ventrocaudal DVN
group f
group x
dorsocaudal LVN
posterior
peripheral SVN
vermis
ventral
tip
of MVN
ventral half of LVN
Iventrolateral DVN
4W
tp
nucleus
-Y
caudal
rostral
-ell
group/x
group f
fastigal
i
V-I
peripheral SVN
dorsal half of LVN
anteriordorsomedial
MVN, except
ventral tip
DVN
vermis
dorsal LVN
ipsilateral
spinal proprioceptive
information
Figure B.3.
dorsorostral DVN
Cerebellar outflow to the vestibular nuclei.
-
205 -
rostroventral LVN
central LVN
medial VST,
ipsilateral
dorsocaudal LVN
lateral VST,
mostly i psilateral
neck and
forelimb
control
trunk
I
hindlimb
laminae VII and VIII of spinal cord
Figure
B.4.
Vestibulospinal tract projections.
lateral,
VST
extensor
muscle
MLF
flexor
muscle
Figure B.S.
Simplified representation of VST and MLF
innervation pattern. All synapses shown are excitatory.
(Grillner, 1969)
-
-
206
except
nuclei,
major
the
(MLF),
medial
ascending
function via the
oculomotor
longitudinal
fasciculus
The
contributes the most to this function.
SVN
nuclei
vestibular
pontomedullary
the
and
nuclei
oculomotor
the
SVN also relays information from
other
to
contribute
DVN,
the
four
the
of
While all
influences on spinal motor systems.
to
reticular
formation.
Each of the major nuclei can be further subdivided into
functionally
differentiated
regions according
nature of their connections with the cerebellum
and
trunk or limb muscles.
controlling neck,
vestibular
nuclei,
segments
Somatotopic mapping
areas
is maintained in the LVN, the cerebellar
the
and reticular
spinal
the
with
connections
the
afferents,
not these regions receive primary vestibular
formation,
to whether or
relaying
to
and to a lesser degree, the MVN and
DVN.
The discovery of such somatotopic mapping combined with
a
knowledge
connections
locomotion.
cortex,
makes
for
strategies
vestibular
of
thalamus,
more
and
basal
CNS
with
structures
such
as
certainly
have
an
ganglia
important influence on motor control,
interaction
during posture and
equilibrium
rostral
likely
of
conception
the
possible
maintaining
While
afferent connections and spinal
the
nature
of
their
more caudal brainstem structures mediating
postural reactions is not known in any detail.
Therefore, in
-
this
study,
207 -
the more automatic processes of posture will be
terms
in
primarily
interpreted
spinal-vestibular-cerebellar
a
more
interpreted
a
phenomenological
from
volitional
explanation
which
sits
must
perspective;
equivalent to postulating a "black box"
structure
Any experimental
interactions.
results requiring
known
of
above the central postural stabilizer
interact
Physiologically,
determined
this is
of relatively unknown
(the vestibular nuclei and associates) and provides
which
be.
with
this
the
postural
translates
interactions
between
commands
stabilizer.
into
experimentally
corticospinal impulses and
vestibular commands, for example.
B.2
Afferent
inflow to the vestibular nuclei
Precht (1974b)
Each type of afferent inflow to the
vestibular
nuclei
usually terminates in a preferred region of each nucleus, and
some types of afferents terminate predominantly
nuclei.
The
vestibular,
main
groups
of
and
specific
afferent fibres are: primary
cerebello-vestibular,
reticulo-vestibular,
in
commissural
afferent inflow arrives indirectly,
via
spino-vestibular,
fibres.
the
Much of the
cerebellum
and
reticular formation.
BeI
Primary yestiUlar afferents
Primary
sensory
information
reaches
the
vestibular
-
nuclei
through
208 -
the
vestibular
vestibulo-cochlear. nerve.
and
acceleration,
receptors;
excite
higher
Low
portion
frequency
gravity
primarily
frequency
angular
stimuli,
excite
utricle
and
saccule
are
linear
otolith
acceleration
stimuli
Otolith input can
be further subdivided into utricular and saccular
The
the
the
semicircular canal receptors.
the
of
believed
to
afferents.
be
primarily
responsible for sensing gravity and linear acceleration.
Most primary
entering
afferent
sprout
terminate
in
functionally
"forelimb"
numerous
the. SVN
collaterals.
and
significant
area,
may
the
Ascending
It
LVN.
rostral
branches
LVN
may be
is
the
to the efferent outflow from the
termination of primary fibres
system.
Some
primary
dorsolateral reticular formation.
rostrolateral
before
in
this
indicate more direct control of the forelimbs by
vestibular
bifurcation
just
these bifurcations,
ventrorostral
that
relative
LVN; the preferential
the
bifurcate
the vestibular nuclear complex;
in turn,
region
fibres
terminate
MVN.
in
fibres
the
Descending branches of the
the
DVN
ventromedial
and
No primary afferents have been shown to
innervate contralateral vestibular nuclei.
Areas of
afferent terminations are not sharply delineated,
are indicated by
reach
distinct
gradients
of
primary
but rather,
afferent
terminal
densities.
The primary vestibular afferents can be subdivided into
-
canal,
and
utricular
number to the ventral LVN.
reach
fibres
saccular
fibres.
A
the
Utricular
rostral
number
small
relatively
and
DVN
fibres
and a smaller
medial DVN,
MVN,
afferents
Canal
MVN.
rostral
caudolateral
the
to
project
saccular
and
SVN
terminate in the
-
209
of
ventral LVN.
Electrophysiological recordings indicate that LVN neurons, in
general, respond primarily to otolith stimulation rather than
While these
to canal stimulation.
predominant
represent
patterns
termination
of
tendencies, some higher order neurons
receive input from
several
saccule and canal.
Both somata and dendrites of many rostral
receptor
LVN neurons are contacted by terminal
afferents,
in
organs,
boutons
for
of
example,
utricular
addition to terminals of cerebello-vestibular
fibres.
Bt4
Response
patterns
of
neurons
vestibular
central
t
primary vestibular afferent input
Precht (1974)
Fernandez and Goldberg (1971,
Goldberg and Fernandez (1971,
Fujita et al. (1968)
Schor (1974)
Both primary
neurons
discharge
and
secondary
spontaneously
Primaries have resting
second
many
1976a,b,c)
1980)
discharges
at
of
vestibular
relatively high rates.
20
per
second,
order neurons fire at rates of 2 to 30 per second.
few secondaries fire at rates up to 90 per second.
origin
sensory
of
resting
discharge
and
A
While the
in primary neurons is unknown,
210 -
-
the resting discharge of
the
secondary
neurons
vestibular
seems to be due to the reflex effect of convergent input from
since
the primaries,
section
after
to
some
ceases
temporarily
of the vestibulo-cochlear nerve.
It recovers
of
because
possibly
degree,
discharge
resting
excitation
reflex
originating from other sources.
Central vestibular neurons that respond
afferent
input
Of the
response.
canals
been
have
accelerations
with
respond
a
new
0.65 deg/secl ,
thresholds.
have
which
canals,
primary
agrees
with
steady,
is
sensation and nystagmus
spontaneously,
a higher threshold of about 5 deg/sec
however,
maintained
neurons
tonic
Kinetic neurons do not discharge
threshold has been reached,
the
significant change
in
frequency
impulse
angular
changing rate of
gradually
a
step
For
horizontal
The threshold for a statistically
level.
and
the
reaches
firing which eventually
in
of
horizontal
the
canals,
most.
the
studied
in the plane
neurons
tonic
of
'pairs
three
"tonic" or "kinetic"
a
either
demonstrate
vestibular
to
.
Once the
kinetic neurons are more
sensitive to further increases of acceleration
than the tonic
functional
importance
for
In addition to the tonic-phasic
dichotomy,
vestibular
neurons,
of
which may be
large
angular accelerations.
neurons
responding
be categorized
to horizontal canal stimulation may also
according to whether the firing rate increases
- 211 -
or
a given direction of angular acceleration.
to
decreases
All primary neurons in the nerve
respond
canal
horizontal
increased
discharge
to
utriculopetal
and
decreased
discharge
to
utriculofugal
(Utriculopetal
motion.
the
of
with
motion,
cupula
branch
endolymph
flow occurs on the right
These
side for clockwise acceleration the horizontal plane.)
as
are
common.
It should be evident that because of the spontaneous
such
discharge,
Secondary
referred
to
neurons
neurons
a
type I units and are the most
bidirectional
demonstrate
response
several
is
other
possible.
types
of
responses as indicated in the following table:
utriculofugal
utriculopetal
type I
+
type II
+
type III
type IV
The
sign
+
+
the
indicates
direction
of
change
in
the
firing rate.
Primary neurons in the nerve branch from
canals
have
horizontal canals.
motion
thresholds
higher
In these
produces increased
the
inverse
of
vertical
than the afferents from the
neurons,
utriculofugal
cupula
discharge and utriculopetal motion
produces decreased discharge.
show
the
this
motion in the vertical plane.
Some secondary
pattern
sensory neurons
in response to angular
-
212 -
In response to sustained acceleration, the firing
of
a
neuron
tonic
is
this
whether
unclear
is
It
maximum.
response, or
The
whether it is due to central neural habituation.
response to an impulse in
sudden
order
firing
primary neurons of the cat show some oscillation in
of
rate
low
a
to
due
oscillation in the peripheral neuron
frequency
a
reaching
after
somewhat
decreases
rate
may
account
for
much
adaptation resulting during maintained
Otolith afferents
also discharge
neurons
vestibular
secondary
stimulation show increased
which
the
In addition, higher
stopping of a rotating platform.
mechanisms
from
resulting
deceleration
of
the
sensory
accelerations.
spontaneously.
to
respond
Most
otolith
discharge during ipsilateral
tilt
during contralateral tilt.
Higher
order neurons exist which show increased or decreased
firing
and
rates
discharge
decreased
analogous to the types of
both directions of tilt,
to
units responding
types of tilt
to
canal
horizontal
c(
+
type /A
-
type i
+
(type
These
neurons are often classified as follows:
ipsilateral
type
stimulation.
c&
not found)
contralateral tilt
+
-
213 -
where sign indicates direction
Most
of
these
of
change
of
firing
rate.
neurons are sensitive to tilt about only one
axis and are mutually orthogonal.
LVN
neurons
stimulation
in
have
the
cat
which
characteristic
respond
response
vestibular
predominantly to
the
afferent
cervico-thoracic
segments.
responses.
The
hindlimb
lumbosacral
segments,
neurons
neurons,
show
which
region,
These
the
are
terminals
forelimb
which
restricted
show
canals,
or
type
project
o(
to
This
reactions
Most of the neurons show graded
responses to varying degrees of tilt.
to tilt.
transiently
caused by convergence
to
projects
Y responses.
mostly type
and will be discussed later.
respond
which
nucleus.
has significance in terms of the observed postural
neurons
otolith
patterns
differ in the hindlimb and forelimb regions of
Primary
to
of higher order
In addition, a few LVN
This could either be
information
from
the
by transient neural response to changes in input
from the otoliths.
Goldberg and Fernandez (1980,
located
the
neurons
SVN
which
vestibular
decrease
exert
afferents.
the
squirrel
monkey)
interposed between the abducens nucleus and
excitatory
influence
upon
primary
When these neurons of the vestibular
efferent system are stimulated,
to
in the
response
the excitatory influence acts
gain
of irregularly discharging
primary neurons to natural vestibular
stimulation.
It
was
214 -
-
proposed
that
dynamic
range
movements,
such a mechanism might function to extend the
which
during
are often very
accelerations
the
head
intended
during
afferents
the
of
large.
Bj5
Commissural connections
The SVN is the source of most of the commissural fibres
to
the
contralateral
The majority of
nuclei.
vestibular
these fibres terminate in the contralateral SVN, but
the vestibular nuclei receive some projections.
fibres
sends a significant number of
DVN,
MVN,
and
vestibplar
the
afferents
of
The DVN also
contralateral
the
Most of these fibres terminate in the
LVN.
ventral aspects of
to
all
contralateral
No
nuclei.
primary
been detected which relay to the
have
contralateral nuclei.
vestibular
vestibular
induced
nuclei
nerve
in
potentials
Recordings of field
by
stimulation
pulse
that
indicate
contralateral
the
both
of
inhibitory
the
and
excitatory influence crosses to the other side with latencies
of several milliseconds.
influenced
primarily
in
II neurons
homolateral
the
II
type
contralateral
neurons
are
labyrinth
via
Mild stimulation of the vestibular nerve
commissural fibres.
results
by
Most of the
inhibition of contralateral type I neurons;
probably
type
opposite side.
I
act
as
neurons,
inhibitory
mediating
Kinetic type I neurons
interneurons
type
upon
the stimuli from the
have
been
shown
to
- 215
-
receive monosynaptic excitation during ipsilateral vestibular
nerve
and
stimulation
neurons is as
inhibition for canal-type
canal
inhibits
stimulation
canal
units; posterior
anterior
units;
canal
and,
horizontal canal
contralateral
inhibits
anterior
well
the
as
This
by
allowed
bidirectionality
stimulation
canal
inhibits contralateral posterior canal units.
as
horizontal
follows:
contralateral
stimulation
during
pattern for crossed
The general
stimulation.
contralateral
inhibition
disynaptic
scheme,
spontaneous
firing, enhances the range of firing rates in response
given
range
ipsilateral
opposite
of
stimuli.
In
each
case,
side is augmented by reinforcing
side.
form
posterior
canal,
from the
input
from
the
sense
opposite
from
whereas both posterior canals and both anterior
other,
canals
a
Note that because of their position, the two
horizontal canals detect rotation in a
each
input
to
pairs
canal
which
which
detect
motion
similarly.
Each
is coplanar with the contralateral anterior
explains
the
utility
of
the
particular
arrangement of crossed inhibition described above.
B.6
Direct spinal afferents
Rubin et al.
Ezure et al.
(1977)
(1978)
Direct spinal afferents to the
relatively
cerebellum.
spinal
sparse
compared
Some of
hindlimb area,
the
to
direct
the
vestibular
nuclei
are
massive input from the
input
originates
in
the
which lies below the caudal end of the
- 216
This direct
itself.)
from the column of Clarke
arise
(DSCT)
(Most dorsal spinocerebellar tract
column of Clarke.
fibres
-
spinal projection is entirely ipsilateral, and terminates
ventrocaudal
on
ventrolateral MVN, groups x and z, with a
DVN,
These areas do
smaller number projecting to dorsocaudal LVN.
not receive primary vestibular afferents.
The direct spinal input is
indirect
areas
cerebellum.
limb,
of
but
Most
EPSP's
the
the. hindlimb
Both
inhibitory.
or
spinal
inhibitory
but
input via the inferior olive and cerebellum
spinal
may be excitatory
forelimb
excitatory,
primarily
and
LVN receive indirect excitatory and
via
input
inferior
the
olive
and
of this input arises from the ipsilateral
have
been
responses
in
detected
to
peripheral stimulation of the contralateral limb.
In order for vestibular sensory input to be useful
postural
control,
a
head-body transformation must occur at
This implies
some level of processing.
information
must
in
by
route
of
the
the
neck
that
nuclei,
vestibular
cerebellum
afferent
vestibular
with
converge
eventually
information either directly
indirectly,
for
or
or
reticular
neck
afferents
formation.
Ezure et al.
(1978) showed
that
the
participate in the vestibulocollic reflex.
of dorsal roots C1 to C4, the
individual
motor
units
was
response
unchanged,
Following section
gain
but
and
the
phase
of
gain
as
- 217
by
measured
the
EMG
-
was
response
The
diminished.
of individual motor unit firing rates was reduced,
component
which suggests that the deafferentation may result
motoneurones
a
indicating
pattern
in
alpha
with
activity,
alpha-gamma coactivation.
of
fewer
Spindle afferents
participating in the reflex.
continued to show modulation in phase
root stimulation indicated that the largest
in
DC
Dorsal
observed
EPSP's
the motoneurones were mediated via polysynaptic pathways;
the identity of these pathways was not
but
determined,
the
medial vestibulospinal tract was considered unlikely.
SCerebellar function in relation to the vestibular nuclei
Pompeiano (1974)
Precht (1974b)
With
function
respect
can
be
to
the
divided
vestibular
nuclei,
cerebellar
into three convenient categories:
vestibular input to the vestibulo-cerebellum, output
vestibulo-cerebellum
to
spino-cerebello-vestibular
the
nuclei,
vestibular
indirect
pathway for
of
the
and
the
information
from the spinal cord to the vestibular nuclei.
B.7.1
General cerebellar physiology
Bell and Dow (1967)
Before reviewing the input and
output
tracts
of
the
cerebellum,
a few points concerning cerebellar physiology are
relevant.
The
somatotopically.
cerebellar
Any
cortex
processing
is
highly
organized
which the cortex performs
-
218 -
spread
of
cerebellum
is
tends to be very localized, as opposed to diffuse
types
ipsilateral, via two basic
fibres.
mossy
of
the
climbing
fibres:
strongly
inferior olive and terminate on Purkinje cells with
excitatory
which
fibres,
ultimately
All
synapses.
either
granule
excite
directly
inhibit
input
other
by
cells
and
can
excite Purkinje cells due to
or
interposed
the granule and Purkinje cells.
Purkinje
cell
which
axons,
are
the
only
efferent
from the cerebellar cortex, synapse primarily on
connections
the cells of the deep cerebellar nuclei, and are
The
mossy
arrives
inhibitory stellate and basket cells which may be
between
and
solely from the
originate
fibres
Climbing
to
input
the
of
Most
activity.
deep
nuclei
also
receive
fibres.
collaterals of climbing and mossy
fire at very high rates,
excitatory
inhibitory.
input
Nuclear
from
neurons
providing a generally tonic efferent
outflow upon which a rate modulation is superimposed.
B.7.2
Input- to the vestibulo-cerebellum
Marini et al.
(1975)
Either primary or secondary
not
numerous
compared
to
other cerebellar input.
vestibular afferents have been
flocculus,
shown
to
terminate
are
Primary
in
the
nodulus, caudal uvula, ventral paraflocculus, and
a small ventral edge of the dentate
secondary
afferents
vestibular
vestibular
afferents
nucleus.
originate
Many
in
group
of
the
x and
-
ventrocaudal MVN.
similar
to
the
Areas of termination in the cerebellum
primary vestibular afferents,
dentate nucleus and also
from
the
219 -
fastigial
form of ipsilateral
including
nucleus.
All
are
excluding the
significant
projections
of these inputs are in the
mossy fibres.
Climbing
fibres
to
the
vestibulo-cerebellum from the inferior olive probably consist
of indirect input from the vestibular nuclei, but the
nature
of such connections has not been determined.
Note that the areas of origin of most of the
secondary
vestibular
afferents
vestibular
afferents.
Because
vestibular
afferents
links
in
secondary
a
the
vestibular
of
receive
this,
pathway
afferent
somatic
not
cerebellum
spinocerebellar
vestibular and
significant
to
do
the
rather
secondary
than
a
true
Convergence
probably
occurs
degree in the vestibulo-cerebellum;
example, group x receives spinal afferents
primary
probably comprise
pathway.
information
so-called
and
to
of
a
as one known
projects
to
the vestibulo-cerebellum.
The inferior olive is probably an important way station
for
convergence
of
spinal
afferents
commands upon the cerebellum.
poor
postural
symmetry
and,
and descending motor
Lesions of the olive result in
during
locomotion,
unsteadiness and an asymmetric hindlimb gait.
that
the
vestibulo-cerebellum
may
function
result in
This
suggests
to
integrate
proprioceptive information and motor commands with vestibular
220
-
-
information in order to correct posture.
The nodulus, in particular, receives both
secondary
and
otolith
-type
bilateral
integration
units
input.
and
representation
with
primary
and
The presence of equal ratios of eX
some
from
oculomotor
Y units
the
and
indicates
vestibular
neck
strong
nuclei.
afferents,
output may serve to regulate head and trunk position.
By
nodular
-
B.Y.
221 -
Output -f the vestibulo-cerebellum
Ito et al. (1968)
Most studies to date confirm the
of
inhibitory
the vestibulo-cerebellar cortex on the vestibular nuclei.
This is consistent with the general principle
cells
exert
that
cerebellar
nuclei;
in
to
vestibulo-cerebellum
respond
Purkinje
to
cells
canal
four
types
of cells found in the vestibular nuclei.
All of the vestibular nuclei
vestibulo-cerebellum,
the
deep
from
horizontal
stimulation with patterns similar to the various
scanty.
the
both cases the Purkinje cells act to
modulate tonic activity of the nuclei.
(I to IV)
Purkinje
inhibition, and suggests that, in some aspects,
the vestibular nuclei are functionally similar
the
influence
although
A general principle of
flocculus
projects
primary vestibular
to
input
from
the
the projection to the LVN is
organization
those
afferents,
receive
and
regions
the
applies
also
nodulus
here:
receiving
and
uvula
project to roughly similar regions to which the contralateral
fastigial
describing
nucleus
also
projects.
fastigio-vestibular
medial
MVN,
section
projections.)
and uvular projections terminate in
and
(See
peripheral
B.7.5
These nodular
SVN,
ventrocaudal DVN, group f, group x.
caudal
A few
fibres also project to dorsocaudal LVN.
The vestibulo-cerebellum seems to be primarily involved
with
oculomotor
function, although it may have significance
-
222 -
Lesions
for postural regulation.
the
in
cortex
of
this
region result in enhancement of vestibular-induced nystagmus,
the
of
appearance
of
Lesions of the nodulus impair otolith influence
habituation.
on the oculomotor
B.7.4
lack
and
nystagmus,
spontaneous
system.
Direct projections .of- t he
cerebellum
spinal
to
the
vestibular nuclei
Ito et al. (1968)
Ito et al.
(1970b)
The anterior vermis of the cerebellum,
mostly
spinal
proprioceptive
which
information,
dorsorostral DVN and dorsocaudal LVN.
The
receives
projects to the
spinal
input
to
the vermis is generally ipsilateral, and the outflow from the
vermis to the vestibular nuclei via the
peduncle
is also ipsilateral.
inferior
cerebellar
Note that the dorsocaudal LVN
to which the anterior vermis projects covers
parts
of
both
cells
from
the forelimb and hindlimb regions of this nucleus.
A
direct
anterior
vermis
inhibitory
to
electrophysiologically,
have
also
detected
path
the
dorsal
LVN
Purkinje
has
been
confirmed
although some investigators claim
excitatory
stimulating anterior vermis.
this
via
The
pathway is 15 to 20 m/sec.
efferents to LVN neurons by
conduction
velocity
of
possibilities
over
Difficulties of stimulation
experiments of cerebellar cortex become evident here,
number
to
since a
complicate the interpretation;
the
-
223 -
presence of basket and stellate cells
inhibit
Purkinje
cells,
the
in
the
cortex
which
presence of Golgi cells which
normally provide recurrent inhibition of parallel fibres, and
possible
antidromic
activation of spinocerebellar
afferents
In
general,
which have collaterals to the fastigial nuclei.
cerebellar
cortex
experiments are difficult to
stimulation
interpret.
Fastigial nuclei
B.7-.5
Ito et al.
(1970a)
An indirect pathway exists from the Purkinje
cells
vermal cortex to the vestibular nuclei via the fastigial
the
nucleus.
fastigial
nuclei.
system,
This route
nucleus
inhibitory
includes
and excitatory
in
synapses
A very specific somatotopic mapping exists for
also.
Vermal
pole
regions
of the fastigial nucleus.
which is the forelimb region,
pole
of
the
fastigial
hindlimb projection.
of
posterior
vermis)
of
lobules
I
this
to
III,
rostral
The vermis of lobules IV and
also projects to the rostral
nucleus,
but more medially than the
Vermis of lobules IX
projects
the
synapses in the vestibular
corresponding to the hindlimb area, p'roject onto the
V,
in
to
the
(hindlinb
region
caudal pole of the
fastigial nucleus just medially to the hindlimb area.
Most of the fastigio-vestibular fibres arising from the
rostral
half
of the fastigial nucleus project ipsilaterally
and those of the caudal half
project
contralaterally.
The
-
crossed
224 -
fastigio-vestibular
fibres
terminate
ventral tip of MVN, ventral half of LVN,
the SVN,
ventrolateral
DVN,
Uncrossed fibres terminate in most of
f, and group x.
group
in
LVN,
the MVN except for the most ventral tip, dorsal half of
and
that
and
crossed
uncrossed
fibres
exception
the
with
regions
This pattern of innervation indicates
DVN.
dorsomedial
of
SVN.
the
different
to
project
Excitatory
fastigio-reticular connections are also numerous.
during
firing rate
contralateral
units are
responses,
also
indicate that most units increase their
tilt
lateral
during
tilt
ipsilateral
(type
found.
indicating
&X).
tilt
and
Curarization
proprioceptive
seem to be velocity sensitive,
to
decays
rate.
during
decrease
A smaller number of 1
does
not
feedback
Many of
cause of the observed response patterns.
to a step tilt;
cats
decerebrate
Recordings from fastigial neurons in
and
alter
X
the
not the
is
the
units
as indicated by their response
the firing rate increases
transiently,
then
a rate which is slightly greater than the resting
No anatomical
evidence of macular or ampullar input to
the fastigial nuclei has yet been found.
The fastigial nuclei also receive input from many other
sources,
mostly in the form of collaterals from climbing and
mossy fibres projecting to the cortex.
spinal
nuclei.
proprioceptive
The
existence
These sources include
afferents and precerebellar reticular
of
fastigio-reticular
positive
- 225
may
loops
feedback
-
the
explain
to
help
rate
high
of
continuous firing of most fastigial units.
Electrical stimulation of the fastigial nucleus
causes
inhibition
in SVN
consistent
of
patterns
excitation
and
neurons which can be independently classified as
disynaptically inhibited by stimulation of the
fastigial
nucleus
and
monosynaptically
contralateral
excited
Type II neurons
ipsilateral fastigial nucleus.
or
Type I neurons are
II by horizontal canal stimulation.
type
I
type
by
are
the
excited
by the contralateral fastigial nucleus, and
monosynaptically
probably function as inhibitory interneurones acting on
type
I cells.
B.8
Reticular formation afferents to the vestibular nuclei
Clendenin, et al. (1974)
The organization of input from the reticular
is
not
known
in
much
is
vestibular
nuclei
projecting
elsewhere.
derived
Some
from
of
of the input to the
Most
detail.
collaterals
the
tracts
collaterals to the vestibular nuclei are
VST.
The
reticular
formation
is
the
the
RST,
as
afferents.
well
axons
of
which
sprout
MLF
and
most probable link
between the vestibular nuclei and more rostral parts
brain,
formation
as acting as an additional
of
the
relay for spinal
-
B.9j
226 -
Efferents from vestibular nuclei
to -reticular
formation
Peterson and Abzug (1975)
All of the four major vestibular nuclei project to
pontomedullary
reticular
formation,
different pattern of termination.
the
LVN
and
reticular
SVN
nucleus
project
and
the
and each nucleus has a
Of these projections, only
to reticular nuclei (the lateral
the
nucleus
reticularis
tegmenti
pontis) which have projections to the cerebellum.
Therefore,
these two reticular nuclei
cerebello-
provide
links
in
a
vestibulo- reticulo- cerebellar loop.
B.10
Influence
of
brainstem
And
cerebellum
on
spinal
The descending motor tracts from the brainstem
include
mechanisms
the
reticulospinal
(VST).
lateral
tract (RST)
The latter can be subdivided
VST
into
two
tracts,
cerebellar
and
These pathways transmit
vestibular
the
processing
to
(-K
motoneurones primarily through interneurones
in
the
cord.
the
(lVST) and the medial VST (mVST), which is part
os the descending MLF.
of
and the vestibulospinal tract
results
and
4Y
spinal
-
B.11
227 -
Lateral vestibulospinal tract (lVST)
Grillner et al. (1970)
Grillner (1969)
Hongo et al. (1975)
Kato and Tanji (1971)
Nyberg-Hansen (1975)
Peterson and Coulter (1977)
Peterson et al. (1978)
Pompeiano (1975)
Wilson and Yoshida (1969)
The
lVST
ipsilaterally.
originates
in
the
LVN
and
descends
This tract is somatotopically organized;
fibres originating in the rostroventral LVN terminate at
cervical
most
the
level (neck and forelimb segments), fibres from the
central are as project to thoracic segments (trunk) and fibres
from
the
dorsocaudal
(hindlimb).
fast,
up
LVN
to lumbosacral
project
The conduction velocity of these fibres is
to
100
m/sec.
medial
motoneurone
cell
group.
border
However,
electrophysiological
monosynapti c contacts exist,
trees
of
O<
which
VII
and
VIII.
are
experiments indicate
implies
and
hip and toe extensor X.
recorded in flexor 4
through
Electrical
monosynaptically excites knee
rarely
bodies
that
dendritic
motoneurones probably extend into lamina VIII.
Polysynapti c effects are mediated
laminae
of
Very few fibres reach
laminae IX where most of the 0( motoneurone cell
located.
very
Most of the fibres at all levels
terminate in laminae VII and VIII of Rexed on the
the
segments
interneurones
in
stimulation of the VST
ankle
extensor
c< ,
but
Disynaptic IPSP's have been
and some hip extensor
C(.
Some of
cells in lamina VIII send axons across the midline,
the
which may
-
effects upon stimulation of the VST.
explain the bilateral
The
LVN,
228 -
which
receives
primarily
otolith
input,
functions to regulate tonic labyrinthine reflexes in the neck
Most of the control it exerts is through VST axons
and body.
or
directly
which
indirectly contact motoneurones of axial
recorded
from
spinal
1
As
a
this
rule,
extensor
shortly
are
responses
CK,
and
can
be
with latencies as short as 3
neurons
msec, with amplitudes of
EPSP's
Monosynaptic
and proximal limb musculature.
4
to
monosynaptic
These
mV.
followed by polysynaptic potentials.
excitation
monosynaptic
disynaptic
recorded in the antagonist
can
IPSP's
flexor
(X.
on
impinges
simultaneously be
This
inhibition
is
probably mediated via Ia interneurones.
Unilateral lesions of the LVN in rats cause
asymmetric
posture with outward splaying of the limbs ipsilateral
to the
the
body,
staggering
gait
and tremor.
tilting
after
lesion, prolonged fits of rolling
The rats lie flattened against
In monkeys, lesions
the ground, unable to lift head or body.
of the caudal LVN result in more severe postural disturbances
than lesions of any other areas
Symptoms include head tilt
lesion.
no
in
elevating
nuclei.
and falling toward the side of the
SVN lesions result in similar symptoms,
difficulty
the
body
descending influence of a tonic nature
much.
vestibular
the
except
that
occurs, suggesting that
is
not
affected
as
In all of these cases, some compensation occurs after
229 -
-
a period of
necessary
recovery.
are
activity
tonic
mechanisms
The
appears that the remaining areas
for
regaining
the
known in detail, but it
not
of
the
nuclei
vestibular
somehow compensate for the lost tonic activity.
tract (mVST)
Medial vestibulospinal
B.12
Grillner et al.
(1971)
The mVST is less prominent than the lateral tract.
the
in
originate
fibres
and
MVN
descend
primarily
ipsilaterally with the MLF, along with fibres of the RST
interstitiospinal
led
spinal
the
cord
to some confusion of terminology in the literature.
The confusion can be partly avoided by noting that
and
and
The fact that the fibres of these
tract.
tracts are so closely associated in much of the
has
The
descending
MLF are not synonomous.
the
mVST
Stimulation of
the MLF at levels above the mid-thoracic segments will result
No fibres of the mVST
in stimulation of all of these tracts'.
have been traced
primary
function
mid-thoracic
below
is
area)
and
Thus,
the
MVN
the
receive
LVN
(but
primary
not
control
the
vestibular
afferents suggests that the neck and forelimb are under
direct
its
probably control of neck and forelimb.
The fact that the forelimb areas of
hindlimb
segments.
more
of the vestibular senses than the hindlimbs.
Conduction velocities of mVST fibres average 63 m/sec.
Neurons in the MVN
neck
"C and
exert
monosynaptic
inhibition
monosynaptic excitation on forelimb flexor
on
O<.
- 230 -
Electrical
stimulation
of
the
labyrinths
also
produces
disynaptic IPSP's in neck 6<, probably through a relay in the
MVN.
See figure B.5 on page 200.
B,13
Caudal vestibulospinal
tract
Peterson and Coulter (1977)
Peterson et al. (1978)
A new projection from
been
has
cord
spinal
the
nuclei
vestibular
named
discovered,
to
the
caudal
the
of
It projects from the caudal poles
vestibulospinal tract.
the MVN and DVN and from the cell group f to the cervical and
lumbar spinal cord.
IVST
and
mVST
This tract is
since
it
projects
ventral and dorsolateral funiculi.
much
anatomically
unlike
the
bilaterally in both the
Conduction velocities are
slower (12 m/sec) than for the 1VST and mVST.
The mode
of action on spinal motoneurones has yet to be determined.
B.14
Reticulospinal tract (RST)
Grillner et al.
About
reticular
half
of
(1971)
the
neurons
in
the
pontomedullary
formation contribute fibres to the RST, especially
neurons
of
the
nucleus
nucleus
reticularis
reticularis
gigantocellularis.
pontis
as
these fibres
far
as
synapse
Conduction velocities
the
on
and
This tract descends
with the mVST in the MLF, but continues beyond
segments
caudalis
the
thoracic
lumbosacral enlargement.
Most of
layers
VII
and
VIII
range from 90 to 130 m/sec,
of
which,
Rexed.
like
-
231 -
the fibres op the VST, is very fast.
stimulation
Electrical
of the caudal pontine reticular formation evokes monosynaptic
EPSPts in flexor
fA
and a few flexor
polysynaptic IPSP's in extensor c(
The 1VST and
the
deduced
been
with
former
as
exceptions
some
includes other tracts in addition to
do
tracts
the
not
below
extend
figure B.5 on page 200,
an
for a
mid-thoracic
limbs
lower
other
See
segments.
of
schematic
mentioned
these
RST,
the
causes
Although the MLF
MLF stimulation.
from
0(.
constitute
The effects of RST activity on the
previously.
have
MLF
of
stimulation
and
activation,
extensor
the
and
Stimulation of the latter causes
antagonist pair, as a rule.
flexor activation,
of
tracts
extensor- c (,
toe
and some hip and
antagonist
the
principle of VST and MLF activation.
B.15
Controll of the Y
motor system
Grillner (1969)
brainstem
The
motoneurones
and
e
is that
and
cerebellum
is
frequently
However,
movements.
separate
connections from o< to
polysynaptic
of
activation
Y have
excitation
blocking of the extrafusal
may
the
in
rule
O(
most
what has been demonstrated
4Y activation is not dependent
that
control
exert
which is largely independent of o( control.
coactivation
intentional
and
can
not
occur.
endplates,
occur.
been
excitation,
0<
upon
Monosynaptic
found,
Following
iY
but
weak
selective
motoneurones
can
-
both
activate
will
stimulation
controlled by the vestibular nuclei by
to control of
of
Y
more
extensor
than
r<
Flexor
Y , which
exaggerated flexor bias ipsilateral to
anterior
the
This depression is
activity.
eventually compensated to some extent.
recover
analogous
mechanisms
especially
cerebellum,
the
depression
cause
0Y' are probably
c( units.
Lesions of
lobe,
Y.
o and
in
cerebellar
Normally,
contractions.
spindle
intrafusal
resulting
stimulation,
cerebellar
by
fired
be
still
232 -
to
contribute an
may
the
tend
lesion.
Beside
abnormal posture, symptoms include hypotonia, pendular tendon
and
reflexes
of
patterns
excitatory
o'
and
the
response to external disturbances of
unchanged,
basically
It
has
cerebellum is to
originating
elsewhere
level
limb
as
a
into
the
activity in
"(
position
O6
tonic
of
postulated
been
act
phasic
remains
although the general level of tonic
activity is lower and the
higher.
lesions,
cerebellar
Following
tremor.
activity
d
is
that one function of the
switch,
either
excitation
channeling
CX
or
Y'
activity.
However, such conclusions seem highly speculative, given
the
of complexity of the strong excitatory and inhibitory
degree
effects of various neurons in the cerebellum.
The previous generalizations concerning supraspinal
control
and
apply
(dynamic
to the static
motoneurones, a
,
Y
since
motoneurones) seem to be controlled quite
-
differently.
In
spontaneously,
decreases,
the decerebrate cat, both
spinalization,
after
but
'
but
remain unchanged.
are affected by supraspinal
activated
from
by
flexor
reflexes,
tonic
dorsal
reflex
C( and
233 -
I
and
fire
rate
the
This implies that
tonic influences,
whereas
root inflow.
afferents,
r
4
and
segmental
However,
exclude the possibility of supraspinal
are
receive input
in
are coactivated.
a
a
flexor
this does not
d control in certain
circumstances.
The
functional
motoneurones
is
significance
that
spindle velocity versus
control
of
the
activation is that
discharge
allow
to
of
C
discharge
rate
and
as
of
the
well
the
Y
II
firing
result
afferents
increases
and
may
as
spindle
of
firing increases primary afferent
velocity;
Ia
4yd
and
adjustment of the muscle
sensitivity,
position
resting
7Y
the
A much simplified summary of
receptors.
response
they
of
the
Y
(Ia)
resting
decrease the
velocity sensitivity of the Ia afferents.
In order to further understand the significance of
Y
system,
it
properties.of the
is
necessary
various
to
somatic
review
the
efferents.
the
functional
Ia
and
II
afferents are present throughout the body of a muscle and are
sensitive to intrafusal
derivatives,
impinging
which
muscle
spindle
length
and
higher
depend on both overall muscle length and
uy activity.
The Ia afferents are
more
sensitive
- 234 -
to
velocity
and
can be selectively activated by vibration;
are
the II afferents are much less affected by vibration and
Ia afferents respond to the
primarily sensitive to position.
onset of stretch with a transient increase
cessation of firing upon release.
and
a
temporary
Muscle spindles respond to
local pressure with sustained firing.
d and
on
the
rs
activation has a number of specific effects
responses
of
Ia
afferents
to
stretch
which are
summarized below:
(1)
With no
K
bias, resting discharge occurs and
the afferent response has a high dynamic
sensitivity to unloading and some sensitivity
to stretch.
(2)
With
d drive, dynamic sensitivity to stretch
increases, but dynamic sensitivity to shortening decreases.
(3)
With
Y drive,
increases,
the resting discharge
rate
and output is linearly related to
length, with little dynamic sensitivity to
either stretch or unloading.
(4)
With both
and
g drive,
output has a high
dynamic sensitivity to lengthening, but not
to shortening.
Other afferents are important in motor performance
reflex regulation,
and
but are not directly modulated by efferent
- 235
-
tendon
Golgi
of
stretch
Thus,
fibres.
series with muscle
or
located in tendons, in
are
afferents,
organ
afferents,
Group Ib
control like the spindle receptors.
tendon
the
acts as a measure of muscle tension, and the group Ib
organs
rates
firing
their
increased
during
afferents
increase
tension.
Ib afferents are relatively insensitive to passive
muscle stretch.
In addition to muscle receptors,
receptors
important
are
also
cutaneous
joint
and
Passive
in spinal reflexes.
joint sense is impaired after selectively blocking the
position sense during active movement is not
receptors,
but
impaired.
Cutaneous
receptors
ipsilateral to the stimulus.
soles
of
afferent
sensations
the
are
excites
which
reflex,
flexion
joint
responsible
fast
flexor
the
for
o(
units
The cutaneous receptors in
the
feet may be especially important in providing
information
for
postural
control.
Differential
of pressure from posterior to anterior and medial
to lateral areas
of
the
soles
of
the
feet
may
provide
substantial cues for body center of gravity determination.
-
236 -
Functional characteristics _Q.
B.16
stretch
reflex
Dufresne et al. (1978)
Ghez and Shinoda (1978)
Herman et al. (1973b)
Jack and Roberts (1978)
Stein and Bawa (1976)
In order to assess the capability of segmental reflexes
to
provide corrective movements, a number of characteristics
are crucial: afferent discharge pattern, neural time
tonic
versus
phasic
motoneurone
discharge,
muscle stiffness (viscoelastic properties).
is
and
delays,
inherent
The purpose here
not to examine the details of particular reflexes, but to
establish general qualities of the segmental
will
indicate
the
manner
in
reflexes
which
which supraspinal control is
likely to be exerted.
The mechanical properties of muscle
contribute
to
"reflex"
ankle
muscles
to
small-amplitude
indicate that most of the torque
mechanical
properties,
neural reflex activity.
quickly
than
significantly
regulation of static posture during
Studies of EMG and
small disturbances.
may
since
torque
reponses
of
sinusoidal displacements
can
phase
be
lag
accounted
for
by
is independent of
Peak torque was developed much
more
could be accounted for by reflex compensation.
In addition, the frequency spectrum of the torque response to
a
not
tendon
tap contains higher frequency components which are
present
sinusoids.
in
This
the
frequency
response
obtained
from
suggests that nonlinear mechanical factors
such as "stiction" act immediately to hinder small movements,
- 237 -
but
as
that
the
stiction
is
the
overcome,
more linear
qualities of the muscle become dominant.
For large postural disturbances, the segmental
is
reflex
to
insufficient
position.
the
correct
latency responses initiated by higher CNS structures
most
the
of
initiates
corrective
Proprioceptively
posture
during
also
account
for
circumstances.
may
corrections
initiated
provide
system
the
upon
depending
ac.tion,
Longer
vestibular
of
can
proprioception
but
instability,
control
descending
powerful
The
action.
corrective
stretch
occur
either
before or after vestibular corrections.
With the aid of a servo-controlled platform which could
effectively alter the gain of ankle feedback from +1
the
to -1, Nashner was able to demonstrate
adaptability
the long latency response due to propriception.
stood on the platform as a
200
in ankle torque occurred
vestibular
feedback.
The subjects
the
platform
only
of
functional stretch
120
300
msec
demonstrated
msec,
reflex
to
later,
due
to
ankle feedback gain was normal,
When
about half of the subjects
developed
of
rotation
of
When ankle feedback was removed (gain=0), changes
occurred.
latency
step
(normal)
which
(FSR).
was
The
a
torque
response
referred to as the
remaining
subjects
torque only after the longer "vestibular" latency.
However, when ankle feedback gain was suddenly changed to -1,
the
subjects
who
responded with the FSR were destabilized.
-
238 -
After several more trials at a gain of -1, the FSR
This
disappeared.
was
interpreted
as
responses
a
for
evidence
mechanism which presets the state of segmental reflexes based
on learning and an expectation of types of disturbances to be
If the prediction is incorrect, then the reflex
encountered.
preconditioning must readapt to the new situation.
D.17
X
-
interaction during reflexes and movement
Allum (1975)
Allum and Budingen (1976)
Bizzi et al. (1978)
Burke and Eklund (1977)
Evarts et al. (1970)
Henneman (1974e,f)
Houk (1974)
Houk and Henneman (1974)
Kots (1976)
Marsden et al. (1977)
Morin and Pierrot-Deseilligny
(1977)
Monosynaptic facilitation of O( motoneurones by spindle
long been known, but the significance of such
has
afferents
debated.
feedback is still
high
not
is
reflex
Allum has shown
develop significant force.
to
enough
The gain bof this
that in subjects instructed to resist movement,
deflection of
the arm is countered by viscoelastic properties for the first
followed by a long latency increase in force
100 msec,
acts
to
fully restore position.
The reflex increase at 100
The
msec is probably due to the segmental reflex.
spindle
and,
feedback
also,
preparation
to
for
may
the
role
be to activate suprasegmental
facilitate
the
descending
which
corresponding
impulses
O
of
centers
units
to follow.
as
This
- 239 -
Yr
theory is supported by the fact that many
large
receptive
fields,
which
would
units have
unsuitable for
seem
precise reflex control via segmental reflexes, but
to a potentiation effect upon 0<
conducive
be
units by means of
Time delays and muscle properties combine
response
disturbances.
velocity
would
spindle feedback.
generalized
sluggish
very
to. corrective
This
can
feedback,
be
which
impulses
compensated
is
to
cause
during
to
an
a
large
extent
by
provided by the Ia afferents.
However, according to experimental
results,
this
does
not
stabilize or damp the response as is commonly proposed, since
the relative gain at the resonant frequency
velocity
It
feedback.
was
suggested
is
that
oscillatory response to brief stretches is,
with
higher
the
damped
in fact,
due
to
such feedback and may be a source of normal tremor.
Both
during
voluntary
corrections
and
movements,
coactivated.
A
contraction
r
allows
bias
the
sensitive
but with proper
more
oC and
which
to
be
c'
postural
are
motoneurones
remain
Ia afferents
during
muscle
in their most
are
ten
times
to small stretches than to larger stretches,
d'
activation, the sensitive
maintained during active movements.
to
automatic
increases
spindles
sensitive, region of operation.
more
and
servo-assisted,
so
disturbances can be corrected.
that
region
can
be
This allows the movement
unexpected
external
240 -
-
Most supraspinal
movement
exert
often
which
kind
of
movement,
4
various
controlled
suppressed.
and
For
are activated
results
Jd^
and
in
reflex
but during movement, segmental
during
example,
are usually
velocity
increased
positions,
is useful to regulate maintained
and
sensitivity
and
the differing requirements
activity.
activation
4)
posture
The advantages of this
are affected.
be appreciated by recognizing
during
control
control of
differential
only the
Sometimes,
can
centers
of
excitability
the
flexor reflex is not desirable and would tend to destabilize.
has
The administration of DOPA
the
determining
during
DOPA
movement.
norepinephrine,
of
patterns
is
been
excited by dorsal root
inactive.
causes
.
regulatory
In
During
afferents,
locomotion,
properties
group
II
depressed,
of
and
of
position.
III
01
of
precursor
are
but
in
4
cd
are reflexly
activation
dL discharge,
extension phase, however,
regulation
system
but is not found
the spinal cat,
flexor
suppression of flexor
and
r
in
relatively
DOPA administration, norepinephrine is
Following
and
released
the
important
an
useful
which is a neurotransmitter at many synapses
neurons.
cord
of
activation
in brain regions controlling locomotion,
spinal
very
and
total
but excites extensor
rapid
flexion
the
flexors
is
must
provide
both
rather
than
important.
The
movement
and
Short latency reflex effects from
(cutaneous
and
joint)
afferents
but long latency reflexes are not affected.
are
This
-
241
-
suggests that group II afferents may be
signaling
reflexes.
more
important
muscle length rather than for facilitating
for
spinal
242 -
-
B,18
Effects 2
on
activity
VST
and
LVN
motor
activity
the
myotatic
initiated elsewhere
Gernandt (1974)
Pompeiano (1975)
The LVN exerts facilitatory influence on
of extension, which is essential for postural tone.
reflexes
effect
This facilitation does not have a continuous
but
reflexes,
segmental
seems
up
by
VST
axons
15
msec,
be
in
by
VST
Local reflex activity inhibits
to 25 msec.
activity
up
the VST impulses must act as
that
indicating
may
reflex
preceded
when
the effect of VST impulses following the reflex
to
the
as a trigger, even
myotatic
strength
gastrocnemius increases in
impulses
the
example,
For
act
the
by
though the impulses transmitted
continuous.
to
on
triggers of local reflex activity, after which the
segmental
activity completely dominates.
stimulation
During high frequency
nerve,
both
and
flexor
extensor
the
of
segmental
reflexes
augmented, but a potent inhibition follows the
the
stimulation,
potent inhibitory
maintained
ampullary
firing
nerve
ipsilaterally
and
contralaterally.
which
lasts
for
is
after-effect
of
thought
reticulospinal
stimulation
limb
several
results
flexion
cessation
minutes.
to
neurons.
in
vestibular
limb
be
due
are
of
This
to
Unilateral
extension
and curvature of the trunk
- 243 -
VST
activity
immediately
When
also
intersegmental
affects
by
precede
stimulation of the brachial
plexus
stimulation
reflexes.
of
the
results in a much
VST,
reduced
lumbosacral ventral root response, which is an intersegmental
proprioceptive coordination reflex ,
is
stimulated
before
the
However,
when the plexus
lumbosacral
VST,
response is enhanced considerably, if
the
ventral
interval
root
between
plexus and VST stimulation is less than 40 msec.
Another example of a
is the inhibition of labrynthine reflexes during
interaction
normally
accompanied
After
discharge.
is
reaction
off-center
maintained
hindlimb reflexes.
a transient hindlimb extensor
by
elimination
replaced
by
head
of
inhibition
and
positions
of
This suggests that the neck afferents may
extensor
C(.
impulses
VST impulses also interact with corticospinal
a
during
facilitation
continuous
this
afferents,
neck
be responsible for inhibiting lumbosacral
in
head
Movements of the
maintained off-center head positions.
are
intersegmental
complex
somewhat
manner which is very dependent on temporal sequencing.
If a volley of VST impulses preceeds an evoked
impulse
by
10 to 15 msec, the corticospinal response in the
corresponding motor nerve is abolished.
preceeds
the
corticospinal
corticospinal
nerve response is enhanced.
impulse
If
by
the
VST
30 msec,
volley
the motor
-
- 244
Descending tracts from the vestibular nuclei also
a
of
degree
Stimulation of
pathways.
terminals, which
monosynaptic
results
in
reflex
stretch
because the depolarization of
the
afferent feedback.
the
spinal afferent
ascending
in
areas
afferent
presynaptic
causes
over
control
the
vestibular
nuclei
of
and Ib
depolarization
L(
Ia
The
inhibition.
presynaptic
is
have
though,
enhanced,
still
outweighs the
of
reduction
Similarly, stimulation of the caudal
fastigial nucleus results in facilitation of contralateral
(
and presynaptic afferent depolarization.
B.19
Flex=
reflex interaction with the vestibular system
Anderson, Berthoz, et al. (1977)
Anderson, Soechting, et al. (1977a)
Anderson, S-oechting, et al. (1977b)
Bruggencate and Lundberg (1974)
Coulter et al. (-1974)
Hinoki and Ushio (1975)
Tokita et al. (1972)
The flexor
pathway
reflex
afferent
(FRA
or
can be influenced by LVN stimulation.
generally excitatory,
the vestibular nuclei,
anterior vermis.
spinoreticular)
The effect is
providing a positive feedback
loop
with inhibitory control exerted
to
by the
The FRA pathway can be subdivided into
two
components: the bilateral ventral flexor reflex tract (bvFRT)
and the ipsilateral forelimb tract (iFT).
activated
by
diffuse
bvFRT neurons
are
flexion of all four limbs, and can be
directly modulated by descending VST activity by monosynaptic
excitation.
The
bvFRT
crosses
at the segmental level and
245 -
-
proceeds to a relay in the lateral
which
efferents
spinocerebellar
tract
ventral
the
afferent
similar
transmits
iFT
All of these
of
those
from
which terminates directly in
(VSCT),
The
cerebellum.
to
analogous
are
the forelimbs to the ipsilateral reticular
from
information
nucleus,
arise to the anterior vermis.
characteristics
the
reticular
nucleus, from which fibres project to the anterior vermis.
Stimulation of the LVN
by
characterized
are
which
in
contralateral
to
LVN stimulation.
conditioning
of
the
the
contralateral
from
receives
input
probably
provide
In addition, reflex
FRA
a
muscle,
major
and
joint
source
tract
facilitates
The FRA tract,
to VST stimulation.
ipsilateral
EPSP's
response
stimuli
noxious
to
flexion
and
extension
ipsilateral
contralateral
reflexes,
crossed
facilitates
skin
which
receptors,
of tonic excitability to
maintain movements evoked by higher CNS centers.
Proprioceptive impulses from the
muscles
lumbar
and
thoracic
to the vestibular system can become strong enough in
symptoms
cases of whiplash injuries and scoliosis, such that
appear
disturbances.
nystagmus,
Such
ability.
symptoms
and nausea.
all
include
with
vertigo,
of
vestibular
spontaneous
In these cases, righting ability may
Procainization of the painful
also be impaired.
decreases
associated
usually
are
which
these
symptoms
These results indicate
the
and
loci
righting
restores
importance
often
of
spinal
246 -
-
input to the brainstem vestibular centers.
B.20
Sources .DLdecerebrate
rijgidiiy.
Modianos and Pfaff (1976)
Schaefer (1974)
Soechting et al. (1977)
Decerebrate
of
means
rigidity has been studied extensively
brainstem mechanisms to this extreme
cat,
decerebrate
a
number
of
provide an interplay of excitatory and inhibitory
on
the
motor
system.
the
remain which
still
sources
In
state.
tonic
a
cerebellar and
of
contribution
the
determining
as
influences
Total deafferentation abolishes this
rigidity, demonstrating the importance of sensory feedback as
a
tonic
source.
However, post-brachial section of the cord
restores rigidity to the forelimbs.
release
the
of
forelimb
spinal
This
results
segments
section,
the
from inhibitory
If, instead of a
influences arising in the lumbosacral area.
post-brachial
from
the complete cerebellum, the
either
anterior vermis, or the fastigial nucleus has been destroyed,
a
similar
difficult
disinhibition
to
interpret,
results.
since
a
latter finding is
This
complex
interplay
of
within
the
Labyrinthine input is also a source of rigidity in
the
inhibitory
and
excitatory
influences
occur
cerebellum itself.
decerebrate
preparation.
Unilateral
vestibulo-cochlear (eighth) nerve causes
ipsilateral
limbs,
but
if
the
section
flaccidity
opposite
nerve
of
in
is
the
the
also
247 -
-
discussed
As
important.
are
influences
inhibitory
crossed
that
indicating
returns,
rigidity
sectioned,
secondary
previously, the return of spontaneous discharge in
is
lesions
nerve
not
the
both
from
arises
inhibition
Crossed
understood.
eighth
after
neurons
vestibular
contralateral vestibular nuclei via commissural fibres and at
to
the segmental level through inhibitory interneurones
nerve
abolished by bilateral eighth
cat.
of the
interruption
are
6f
and
motoneurones
to restore tonic
necessary
is
by
influences.
labyrinthine
by
cerebellum
the
Therefore, the
loop.
directly
affected
Apparently,
)
altered
significantly
not
are
effects
These
rigidity
section,
abolished in the decerebellate decerebrate
are
opisthotonus
rigidity is not
decerebrate
Although
limb..
contralateral
the
nerve
activity to descending pathways after bilateral eighth
section.
B.21
Eighth nerve section in
Precht (1974a)
Putkonen et al.
intact animals
(1977)
Section of the eighth nerve in otherwise intact animals
(including
humans,
transient
severe
weakness
of
dogs,
disturbance
extensor
spontaneous head turning
nystagmus,
tremor,
tone
toward
pigs) results in a
guinea
cats,
including
equilibrium,
of
of
the
the
side
and staggering gait.
to several weeks, depending upon the
ipsilateral
of
the
limbs,
lesion,
After several days
species,
most
of
the
- 248 -
symptoms
diminish
markedly, indicating neural compensation.
A number of sources may be responsible for such compensation.
proprioceptive,
Tactile,
and
feedback
visual
assume much
greater importance for maintaining balance.
Electrophysiological
vestibular
affected
recordings
nuclei
of
units
that
indicate
in
the
spontaneous
discharge, which is drastically reduced immediately following
the
lesion,
recovers
to
a significant degree.
During the
initial post-operative stage, type II neurons ipsilateral
the
lesion
are not affected,
since they receive
from the contralateral vestibular nuclei.
to
horizontal
to the lesion.
type.
In
Type
their input
I
responses
canal stimulation cannot be found ipsilateral
Type I neurons are normally the
most
neurons,
spontaneous
the
lesion.
which
further
decreases
discharge in vestibular
Eventually
though,
spontaneous discharge.
Whether
spontaneous
or
mechanism,
nuclei
type
this
whether
possibility
seems
unlikely,
by
type
the general level of
I
is
ipsilateral
to
neurons regain a
due
to
a
truly
other afferent channels
sprout collaterals to provide excitation is not
latter
common
to the removal of the primary source of
addition
excitatory input, many type I neurons are inhibited
II
to
since
known.
lesions
The
of the
reticular formation do not prevent the resting discharge from
returning.
249 -
-
B.22
Fastigial1 lesions and atonia
Unilateral
destruction
of
the
fastigial
nucleus,
similar to the effect of LVN lesions, abolishes extensor tone
Only
ipsilaterally, but increases tone contralaterally.
the
rostral pole of the fastigial nucleus needs to be lesioned in
After a
order to observe this effect.
symmetric
lesion
of
the
contralateral fastigial nucleus, rigidity returns due to
the
removal
of
crossed
vestibular nuclei
A
influences
called
lesion
of
the
crossed
fastigial
atonia
caudal
fastigial
nucleus.
symptoms are analogous to rostral fastigial
that
the
contralateral
ipsilateral
crossed
extensors
extensors are facilitated
activity
the
due
influences
in
the
are
occurs
lesions,
are
Following fastigial
inhibition.
crossed inhibitory
tonic
at
and the spinal cord.
syndrome
following
inhibitory
atonic
to
The
except
and
release
the
from
lesions in general,
sufficient
disfacilitated side.
to
suppress
The fastigial
nuclei in the normal person probably function to balance
and
adjust interlimb inhibitory influences.
Corticobulbar Purkinje axons, in their
inferior
cerebellar
the fastigial nuclei.
anterior
increases,
vermis,
peduncle,
to
the
pass just rostrolaterally to
If this bundle,
which arises from
the
is interrupted, ipsilateral extensor tonus
due to release of tonic inhibition.
tonus decreases
course
due to an increase of crossed
Contralateral
inhibition.
-
B.23
250 -
Brainstem and cerebellar function during locomotion
Grillner (1975)
Hayes and Clarke (1978)
Kornhuber (1974b)
Shik and Orlovsky (1976)
Soechting et al. (1978)
Stein (1974)
Thach (1978)
it
movements,
instructive
is
examine
to
cat
locomotion
is
Supraspinal
movement,
it
is
apparent
contained
control
and once
it
that
in
the
the
of
of
the
basic. pattern
for
spinal
provides
the
has
started,
been
role
the
From studies
descending pathways ,during locomotion.
spinal
during
control
To understand the issue of equilibrium
cord
itself.
to
initiate
trigger
the
supraspinal
controllers provide additional tonic and phasic excitation to
maintain the movement and make
fine
adjustments
in
force.
They do not modify the basic pattern of motion, including the
timing and
interlimb
include
pathways
coordination.
the
The
descending
corticbospinal,
reticoluspinal (RST) and vestibulospinal (VST)
motor
rubrospinal,
tracts.
The VST, for example, seems to exert descending control
of
each
limb only during the extension phase.
This fact is
consistent with results of electrical stimulation of the
in
spinal reflexes, which depend upon the timing relative to
on-going
reflex
activity.
Speed
activity
of
and
locomotion
descending
corticospinal
is controlled by the power
developed by the limbs during the stance phase of each
and
VST
it
is
reasonable
that
equilibrium
limb,
should be also be
251
-
-
controlled by fine adjustment of this power during the stance
phase.
The
bipedal locomotion of humans can be viewed as a
continuously unstable process, during
in
fall
one
direction
is
which
a
tendency
by corrective
countered
developed by the extended leg such that the flexed
fall
at
an
optimal pivot point.
to
forces
leg
will
Large errors must also be
accompanied by active participation of the flexed leg and the
arms.
The RST, which.is in many respects, the
the
complement
of
VST, exerts facilitatory influence on flexion during the
flexion phase.
Electrical stimulation of this
extension
no effect.
has
phasic
control
are read, in a sense,
Actually,
phasic
theoretically
be
sufficient
to
since the descending control commands
only
during
activity
during locomotion,
during
Therefore, continuous activity in
both the RST and VST might
exert
tract
but
does
such
the
appropriate
occur
phasic
in
phase.
the RST and VST
activity
requires
an
intact cerebellum.
The cerebellum is also necessary to permit
of
a
movement
without
oscillation.
termination
This is achieved by a
precisely timed contraction of antagonist muscle at
of
are
a
movement.
suppressed
coordination,
the
end
As a movement is initiated, the antagonists
before
together
the
agonists
contract.
Such
with the observed phasic activity of
the RST and VST, suggests that both ascending
proprioceptive
- 252
-
cues and corticospinal impulses may be used by the cerebellum
to control a switch which selectively activates the
RST
during
the
appropriate phases of movement.
VST
and
During the
initiation of locomotion, the corticospinal, rubrospinal, and
reticulospinal tracts are all activated to facilitate flexion
of the starting limb, while VST activity is suppressed.
-
253 -
- 254
-
APPENDIX C
EMG PROCESSOR SCHEMATIC AND SPECIFICATIONS
For a schematic of one channel of
the
EMG
processor,
see figure C.1 on page 250. All of the operational amplifiers
are type A47 4 1 .
Because full wave rectification is a nonlinear process,
the band-pass stage preceding the rectifier and the averaging
filter following the rectifier are analyzed separately.
The
gain,
low
pass,
high
pass
stages
are
by linear transfer functions, which are listed
characterized
The overall magnitude function for
in Table C.1 on page 251.
the
and
band-passing
preceding
rectification
is
plotted
in
figure C.2 on page 252.
Following rectification,
conditions
the
signal
computer at 500 Hz.
averager
the
is
averager
Butterworth
comparison.
a third order averaging filter
prior
The overall
to output for sampling by the
transfer
also listed in table C.1.
is
plotted
filter
step
in
of
the
The step response of
figure C.3
responses
function
are
on
also
page 252.
shown
for
eZ,
~
~
~
2
91
~c
IraLg
5
~I
13?cR
47
--
+
I
~~313
SO
33K
HR619
(D
l
C:
0(
C)t
*
~
t
0
-ISO
'417I
256
-
Table C.1
Transfer functions for individual stages
of the EMG proces
a-9eg
e
e...
*t*ee
In p0 .eutg ai n00est a0e0
Hc(s) =
00
0 0000
*e..
<
s
0a
S+ 30
s+
*0
.
e*
.
e0..
.
*e
oe
a0a00
0*000
e
000g......a00
aa0
9
<10
0e~
0*00*
040e.
~
g
ee c .
Low pass
-3.3 (a')_
-,/R -I 1 C2C3
,
(s) -e
-')C
0g 0e
S2 +
e
e0
..
..
.
l'+3.(0
a
C*O
o
g 0000
9*0*
0000
0000
High pass
+- Ic+
Ks 2 +
RSC5
s0
+
O.25 sZ 4.
i
9
+ ± 9\
\Cs
k
R13Ci
+
- C
C4
0. Y5
P' 7
High pass filtering due to C
S
PC''
S
,
S+
ee ee ee c ge cee g ... e
S?
e
e
e
ee
e
e
e
e
g
00g
g
c
e
Averaging filter
*r
C
C 7 =CS=C,
Rmr\14 RI-S
CS
IL RIq R 17 R
15 R 17 k1
R
\Z
PAV
R R,
IR 12
Rif , R ao, C '3- ( R C-s+
.2T S
+
12L?
S3 5A + 12T'S'+ 60~Ts
z R]
LR1 R, 1 R
+
~VerQ
t-
) 2.o
1)
~'
"
+L >S
0s
sec
- ad
040 0 0
. .
0*
-
257 -
200
500
1.0.
0.10,I
2o
Figure C.2.
/6C
So
IcCo
Zeo0
(H7.)
Magnitude plot for bandpass stage of EMG
processor
1.0 -
.-- averaging filter
,o o 1 st order Butterworth
3rd order
---
Butterworth
/0 racc
=
0
.%
C)
0
C)
0
.0,
'*0 N
0~~~
-
0
C
0
Figure C.3.
1C4
ZC
30G
10
5CC
(Hz-)
Step response of third order averager
compared to Butterworth filters.
- 258
-
APPENDIX D
STATISTICAL ANALYSIS PROCEDURE
Assuming that some measure,
data
X ,6,
is obtained
from
the
for a given subject, block, and drop test, this measure
can be modelled as a sum of components
contributed
each
by
source of variation:
++
(X
-
=total
deviation from mean for
entire experiment
'.)
X
variation due to differences
between individuals
=variation due to differences between
blocks (effect of testing sequence)
(
Kvariation
due to parameters of
the fall
+
)
block by column
interaction term
= remainder term
Since each block uses a different group of
r x c,
and
r x b x c
interactions
subjects,
b x, r,
are meaningless for the
-
split plot design.
Thus,
259 -
the corresponding
terms are
absent
from the model.
Each component of the additive model gives
estimate.
variance
vanish.)
The
(All
formulas
rise
to
a
cross product terms can be shown to
for
these
are
estimates
in
given
Table D.1 on page 256 (McNemar, 1969). The expected values of
these estimates are also given.
In order to prevent spurious detection of "significant"
F-test
values,
those
F-tests
which
are
suspected
to be
significant must be chosen beforehand, along with appropriate
levels
of
significance.
posteriori must'attain a
Any
higher
such
level
tests
of
performed
significance
a
to
compensate for the fact that among a large number of F-tests,
several tests are more
likely
to
exceed
the
significance
level even though the null hypothesis may be true.
If N F-tests are performed
collected,
after
the
data
has
been
and a level of significance, P, , is desired, then
the level of significance, P, at which each
individual
must be performed must be determined as follows:
test
- 260 -
Prob
m or more tests out of N
null
will exceed the sigificance
hypothesis
level
is true
1
1)
l(=O
for
nk
'.
s
Therefore, the criterion for P becomes:
Or
where P
1-
P)
(101(o
is the maximum significance level
For example, if N= 20 and
o 1P)
l
P
(O.A)
tolerated.
0.05,
. CO
.0it
0 0005
This last constraint requires choice of a standard P value of
0.001.
From the expected values of the variance
Table
are:
D.1,
it
can be seen that
estimates
in
the relevant F-test ratios
-
261
Variance estimates for main and interaction
Table D.l.
terms of three way split plot experimental design.
6*
*
*
9.
.
.
.
.
estimate
00
S
.
.
.
..
* 00
6
0
0 0
000 0060. .
*00
sum of squares
#deg. of freedom
variance
60
0X
.00O60
0 0
90
E £s]=expected
value
..
0.
005070
00
0*774*0
.
expanded formula
for computation
. *.-..
y
.
-L
0
0
0 0
0a00a04
0000600.
.
0....
.00
0
60
0
90
.
0*00009*0*000*00*
.
000*0000*4
0 00
.*0
sIc-
.0.*0
0w0
4
a00
000a
a*0 4*0
.
a0
.
.
a..
..
9
*0
S',
R13
0
.
.
0*
.
+
arc
C-1
--
- o
c-I
C
b
--L
RC
1"
C + R13C
Ls *
S 6
6
a
2A-
+
. (6-)(0-)
e
'
+
-
bb
cc
Y- 6
P
00000000
*
Definitions:
0
*
9
0
a
0
6
0
0
RC0
0
a
0
6
4
0
a
0
0
a
a
0
a
a
0
e+c.
S, , = (ZT X,6,
2
A,6
A.
*rC
S-L
6 C 4 -C.
2.
.
-)(C-)
(
-~
060
.C
a0..
-c
2-I
00*00a0
C
'3- "
RC
*.a.0.0
0
a004*0
true individual variance
error
.= - variance due to non-zero rxc interaction
= measurement
=
(A .6.
- .P., *
Ac= ( p.. - .,
A .- = (,p.6c - '.
ix= true mean
)
)
-/L.C
+...
)
-
262 -
which tests for significant differences
due to testing sequence
differences due
to drop
test parameters
b)
!
SK
differences due to b x c interaction
(testing sequence by drop test)
If results of a particular F-test indicate insignificant main
block
effects and insignificant block by column interaction,
then columns may be averaged over all
then
and
columns
F-tests
significant
may
and
subjects,
be averaged over all blocks and subjects,
performed
block
blocks
between
columns.
However,
since
(testing sequence) and block by drop test
interactions are likely, column
interpreted with caution.
average
F-tests
should
be
- 263 -
BIBLIOGRAPHY
Aggashyan, R.V. and Pal'tsev, Y.I. (1975).
Reproduction of
certain special aspects of the dynamics of the maintenance
of the vertical posture by man using a mathematical model.
Biofizika 2_Q:137-142.
Allen, G.R. (1977).
Human tolerance of repeated shocks.
In Proceedings of the European Symposium on Life Sciences
Research in Space, Cologne-Porz, Germany, 2i4-26 May 1977,
ESA SP-130, pp.343-349.
Allum,,* J.H.J. (1975).
Responses to load disturbances in
human shoulder muscles: the hypothesis that one component
is a pulse test information signal.
Exp. Brain Res.
22:307-326.
Responses to load
Allum, J.H.J. and BUdingen, H.J. (1976).
disturbances in human leg muscles: the organization of long
latency reflexes. Presented at the Third International
Symposium on Motor Control, Bulgaria, May 1976.
Allum, J.H.J., Graf, W., Dichgans, J., and Schmidt, C.L.
Visual-vestibular interactions in the vestibular
(1976).
nuclei of the goldfish. Exp. Brain Res. 25.:463-485.
Anderson, J.H., Berthoz, A., Soechting, J.F., and Terzuolo,
Motor output to deafferented forelimb extensors
C.A. (1977).
in the decerebrate cat during natural vestibular stimulation.
Brain Res. 122:150-153.
Anderson, J.H., Soechting, J.F., and Terzuolo, C.A. (1977a).
Dynamic relations between natural vestibular inputs and
activity of forelimb extensor muscles in the decerebrate
cat. I. Motor output during sinusoidal linear accelerations.
Brain Res. 120:1-15.
Anderson, J.H., Soechting, J.F., and Terzuolo, C.A. (1977b).
Dynamic relations between natural vestibular inputs and
activity of forelimb extensor muscles in the decerebrate cat.
II. Motor output during rotations in the horizontal plane.
Brain Res. 120:17-33.
Anderson, M.E., Yoshida, M., and Wilson, V.J. (1971).
Influence of superior colliculus on cat neck motoneurones.
J. Neurophysiol. 34:898-907.
Ansorge, K. and Grusser-Cornehls, U. (1977).
Visual and
visual-vestibular responses in frog cerebellar neurons.
Exp.
Brain Res. 29:445-465.
-
264 -
Basmajian, J.V. (1967).
Muscles Alive, Their Functions
Revealed by Electroyograpy. Williams and Wilkens,
Baltimore.
Bell, C.C. and Dow, R.S. (1967).
Cerebellar circuitry.
In Neurosciences Research Symposium Summaries, vol. 2, F.O.
Schmitt et al. (Eds.), MIT Press, Cambridge, Mass.,
pp.515-616.
Berthoz, A., Pavard, B., and Young, L.R. (1975).
Perception
of linear horizontal self-motion induced by peripheral vision
(linearvection). Basic characteristics and visual-vestibular
interactions. Exp. Brain Res. 23:471-489.
Bizzi, E., Dev, P., Morasso, P., and Polit, A. (1978).
Effect of load disturbances during centrally initiated movements. J. Neurophysiol. 41:542-556.
Brandt, T., Bles, W., Kapteyn, T.S., and Arnold, F. (1977).
Height vertigo: a distance vertigo through visual destabilization.
11th World Congress of Neurology, Amsterdam,
11-16 Sept. 1977, Excerpta Medica International Congress
Series No. 427 (abstract).
Brandt, T., Dichgans, J., and Koenig, E. (1973).
Differential effects of central versus peripheral vision on egocentric and exocentric motion perception.
Exp. Brain Res.
16:476-491.
Brodal, A. (1974).
Anatomy of the vestibular nuclei and
their connections. In Handbook of Sensory Physiology, vol. 6
(Vestibular System), part 1, H.H. Kornhuber (Ed.), SpringerVerlag, Berlin, pp.239-352.
Bruggencate, G. ten and Lundberg, A. (1974).
Facilitatory
interaction in transmission to motoneurones from vestibulospinal fibres and contralateral primary afferents. Exp.
Brain Res..12:248-270.
Burke, D. and Eklund,
man duringistanding.
Chu, W.H.N.
(1976).
G. (1977).
Muscle spindle activity in
Acta Physiol. Scand. 100:187-199.
Dynamic response of human linearvection.
Master's thesis, MIT.
Clark, F.J., Horch, K.W., Bach, S.M., and Larson, G.F.
(1979).
Contributions of cutaneous and joint receptors to
static knee-position sense in man. J. Neurophysiol.
42:877-888.
-
265 -
Clendenin, M., Ekerot, C.F., Oscarsson, 0., and Rosen, I.
(1974).
Functional organization of two spinoreticular paths
relayed through the lateral reticular nucleus in the cat.
Brain Res. _69:140-143.
Cochran, W.G. and Cox, G.M. (1950).
Wiley, Ney York.
Experimental Designs,
Coulter, J.D., Mergner, T., and Pompeiano, 0. (1974).
Macular influences on ascending spinoreticular neurons located in the cervical cord.
Brain Res. a2:322-327.
Daunton, N. and Thomsen, D. (1979).
Visual modulation of
otolith-dependent units in cat vestibular nuclei.
Exp.
Brain Res. 3:173-176.
Dichgans, J. (1975).
Spinal afferences to the oculomotor
system: physiological and clinical aspects. In Basic Mechanisms of Ocular Motility and Their Clinica-1 Implications,
G. Lennerstrand and P. Bach-y-Rita (Eds.), Pergamon Press,
Oxford and New York.
Dichgans, J., Schmidt, C.L., and Graf, W. (1973).
Visual
input improves the speedometer function of the vestibular
nuclei in the goldfish. Exp. Brain Res. 18:319-322.
Dufresne, J.R., Soechting, J.F., and Terzuolo, C.A. (1978).
Electromyographic response to pseudo-random torque disturbances of human forearm position. Neuroscience 1:1213-1226.
Evarts, E.V. (1970a).
Activity of ventralis lateralis
neurons prior to movement in the monkey. Physiologist
13:191.
Evarts, E.V. et al. (1970b).
Central control of movement.
In Neurosciences Research Symposium Summaries, vol. 6, F.O.
Schmitt et al. (Eds.), MIT Press, Cambridge, Mass., pp.1-170.
Ezure, K., Sasaki, S., Uchino, Y., and Wilson, V.J. (1978).
Frequency-response analysis of vestibular-induced neck reflex
in cat.
II. Functional significance of cervical afferents
and polysynaptic descending pathways.
J. Neurophysiol.
41.:459-471.
Fernandez, C. and Goldberg, J.M. (1976a).
Physiology of peripheral neurons innervating otolith organs of the squirrel
monkey. I. Response to static tilts and to long-duration
centrifugal force. J. Neurophysiol. 39:970-984.
Fernandez, C. and Goldberg, J.M. (1976b).
Physiology of peripheral neurons innervating otolith organs of the squirrel
monkey. II. Directional selectivity and force-response
relations.
J. Neurophysiol. 3.:985-995.
-
266 -
Fernandez, C. and Goldberg, J.M. (1976c).
Physiology of peripheral neurons innervating otolith organs of the squirrel
monkey:
III. Response dynamics. J. Neurophysiol.
-12:996-1008.
Fernandez, C. and Goldberg, J.M. (1971).
Physiology of peripheral neurons innervating semicircular canals of the
squirrel monkey.
II. Response to sinusoidal stimulation and
dynamics of peripheral vestibular system. J. Neurophysiol.
34:661-675.
Vestibular ataxia and its measurement
Fregly, A.R. (1974).
in man. In Handbook of Sensory Physiology, vol. 6 (Vestibular System), part 2, H.H. Kornhuber (Ed.),
Springer-Verlag,
Berlin, pp.321-360.
ActivFujita, Y., Rosenberg, J., and'Segundo, J.P. (1968).
ity of cells in the lateral vestibular nucleus as a function
of head position.
J. Physiol. (London) 196:1-18.
Fukuda, T. (1976).
Postural behavior and motion sickness.
Acta Otolaryngol. 81:237-241.
Garland, H. (1972).
A state variable averaging filter for
electromyogram processing. Med. Biol. Eng. 10:559-560.
Gernandt, B.E. (1974).
Vestibulospinal mechanisms.
in Handbook of Sensory Physiology, vol. 6 (Vestibular
System), part 1, H.H. Kornhuber (Ed.), Springer-Verlag,
Berlin, pp.541-564.
Spinal mechanisms of the
Ghez, C. and Shinoda, Y. (1978).
functional stretch reflex. Exp. Brain Res. 32:55-68.
Goldberg, J.M. and Fernandez, C. (1971).
Physiology of peripheral neurons innervating semicircular canals of the
squirrel monkey. I. Resting discharge and response to constant angular accelerations.
J. Neurophysiol. 34:635-660.
Goldberg, J.M. and Fernandez, C. (1980).
Efferent vestibular
system in the squirrel monkey: anatomical location and
J. Neurophysiol.
influence on afferent activity.
A3:986-1025.
Gray, H. (1973).
C.M. Goss (Ed.),
Anatomy of the Human Body.
Lea and Fibiger, Philadelphia, pp.487-526.
Graybiel, A. (1974).
Measurement of otolith function in man.
In Handbook of Sensory Physiology, vol. 6 (Vestibular
System), part 2, H.H. Kornhuber (Ed.), Springer-Verlag,
Berlin, pp.233-266.
-
267 -
Greenwood, R. and Hopkins, A. (1976a). Muscle responses
during sudden falls in man. J. Physiol. (London) 254:507-
518.
Greenwood, R. and Hopkins, A. (1976b). Landing from an unexpected fall and a voluntary step. Brain .9.2:375-386.
Supraspinal and segmental control of
Grillner, S. (1969).
Acta Physiol.
static and dynamic Y-motoneurones in the cat.
Scand. SulPpl. 327:1-34.
Locomotion in vertebrates: central
Grillner, S. (1975).
mechanisms and reflex interactions. Physiol. Rev. 55:247-
303.
The vestibuloGrillner, S., Hongo, T., and Lund, S. (1970).
Effects on alpha-motoneurones in the lumbospinal tract.
sacral spinal cord in the cat. Exp. Brain Res. jiQ:94-120.
Convergent
Grillner, S., Hongo, T., and Lund, S. (1971).
effects on alpha motoneurones from the vestibulospinal tract
and a pathway descending in the medial longitudinal fasciculus.
Exp. Brain Res. 1.2:457-479.
Facilitation of late
Hayes, K.C. and Clarke, A.M. (1978).
reflexes in humans during the preparatory period of volunBrain Res. 153:176-182.
tary movement.
CharacterisHeld, R., Dichgans, J., and Bauer, J. (1974).
orientation.
spatial
influencing
tics of moving visual scenes
Vision Res. 14:1-9.
Vestibular
Henn, V., Young, L.R., and Finley, C.' (1974).
nucleus units in alert monkeys are also influenced by moving
Brain Res. 71:144-149.
visual fields.
Henneman, E.
logy, vol.1,
pp.722-74 6 .
In Medical PhysioThe cerebellum.
(1974a).
V.B. Mountcastle (Ed.), Mosby, St. Louis,
Motor functions of the brainstem and
Henneman, E. (1974b).
basal ganglia. In Medical Physiology, vol.1, V.B. Mountcastle (Ed.), Mosby, St. Louis, pp.668-703.
Motor functions of the cerebral
Henneman, E. (1974c).
In Medical Physiology, vol.1, V.B. Mountcastle
cortex.
(Ed.), Mosby, St. Louis, pp.747-779.
Henneman, E. (1974d).
Organization of the spinal cord.
In. Medical Physiology, vol.1, V.B. Mountcastle (Ed.), Mosby,
St. Louis, pp.636-650.
-
268 -
Henneman, E. (1974e).
Peripheral mechanisms involved in the
control of muscle. In Medical Physiology, vol.1, V.B.
Mountcastle (Ed.), Mosby, St. Louis, pp.617-635.
Spinal reflexes in the control of
Henneman, E. (1974f).
In
Medical
Physiology,
vol.1, V.B. Mountcastle
movement.
6
6
6
(Ed.), Mosby, St. Louis, pp. 517.
Herman, R., Cook, T., Cozzens, B., and Freedman, W. (1973a).
Control of postural reactions in man. In Control of Posture
and Locomotion, R.B. Stein et al. (Eds.), Plenum Press, New
York, pp.363-385.
Herman, R., Freedman, W., Monster, A.W., and Tamai, Y.
A systematic analysis of myotatic reflex activity
(1973b).
in human spastic muscle. In New Developments in Electromyography and Clinical Neurophysiology, J.E. Desmedt (Ed.),
S. Karger, Basel and New York, pp.55 6 -57 8 .
Hershler, C. and Milner, M. (1978).
An optimality criterion
for processing electromyographic (EMG) signals relating to
human locomotion.
IEEE Trans. Biomed. Eng. BME-25:413-420.
Hinoki, M. and Ushio, N. (1975).
Lumbomuscular proprioceptive reflexes in body equilibrium. Acta Otolaryngol. Sugpl.
330:197-210.
Myoelectric prosthesis control: optimal
Hogan, N. (1976).
estimation applied to EMG and the cybernetic considerations
for its
use in a man-machine interface.
Ph.D. thesis, MIT.
Hongo, T., Kudo, N., and Tanaka, R. (1975).
The vestibulospinal tract:
crossed and uncrossed effects on hindlimb
motoneurones in the cat. Exp. Brain Res. 24:37-55.
Hood, J.D., and Leech, J. (1974).
The significance of
peripheral vision in the perception of movement. Acta
Otolaryngol. 71:72-79.
Hosman, R.J.A.W. and Vaart, J.C. van der (1978).
Vestibular
models and thresholds of motion perception.
Results of tests
in a flight simulator. Delft Univ. of Technology, Dept. of
Aerospace Engineering Report LR-265, Delft, the Netherlands,
April 1978.
Houk, J. (1974).
Feedback control of muscle: a synthesis of
the peripheral mechanisms. In Medical Physiology, vol.1,
V.B. Mountcastle (Ed.), Mosby, St. Louis, pp.668-677.
Houk, J. and Henneman, E. (1974).
Feedback control of
muscle: introductory concepts. In Medical Physiology, vol.1,
V.B. Mountcastle (Ed.), Mosby, St. Louis, pp.608-616.
- 269 -
Huang, J.K. and Young, L.R. (1980).
Sensation of rotation
about a vertical axis with fixed visual field in different
illuminations and in the dark. Accepted for publication in
Exp. Brain Res.
Ito, M., Kawai, N., and Udo, M. (1968).
The origin of cereb.ellar-induced inhibition of Deiter's neurones.
III. Localization of the inhibitory zone.
Exp. Brain Res. 1:310-320.
Ito, M., Udo, M., Mano, N., Kawai, N., (1970a).
Synaptic
action of the fastigiobulbar impulses upon neurones in the
medullary reticular formation and vestibular nuclei.
Exp.
Brain Res. 11:29-47.
Ito, M., Yoshida, M., Obata, K., Kawai, N., and Udo, M.
(1970b).
Inhibitory control of intracerebellar nuclei by the
Purkinje cell axons. Exp. Brain Res. 10:64-80.
Jack, J.J.B. and Roberts, R.C. (1978).
The role of muscle
spindle afferents in stretch and vibration reflexes of the
soleus muscle of the decerebrate cat.
Brain Res. 146:366372.
Kapteyn, T.S. and Wit, G. de (1972).
Posturography as an
auxiliary in vestibular investigation.
Acta Otolaryngol.
l3:104-11.
Kato, M. and Tanji, J. (1971).
The effects of electrical
stimulation of Deiter's nucleus upon hindlimb -motoneurones
in the cat.
Brain Res. 30:385-395.
Kaveh, M., Bruzzone, S., and Torres, F. (1978).
for the estimation of averaged evoked responses.
Systems, Man, and Cybernetics SMC-8:414-417.
A new method
IEEE Trans.
Keller, E.L. and Kamath, B.Y. (1975).
Characteristics of
head rotation and eye movement-related neurones in alert
monkey vestibular nucleus.
Brain Res. I1.0:182-187.
Keller, E.L., and Precht, W. (1978).
Persistence of visual
response in vestibular nucleus neurons in cerebellectomized
cat.
Exp. Brain Res. 32:591-594.
Keller, E.L., and Precht, W. (1979a).
Adaptive modification
of central vestibular neurons in response to visual stimulation through reversing prisms.
J. Neurophysiol.
42:896-911.
Keller, E.L., and Precht, W. (1979b).
Visual-vestibular
responses in vestibular nuclear neurones in the intact and
cerebellectomized, alert cat.
Neuroscience !:1599-1613.
-
270
-
Joint receptors do not provide a
Kelso, J.A.S. (1978).
satisfactory basis for motor timing and positioning.
Psychol. Rev. 85: 474-481.
Nystagmus and related phenomenon
Kornhuber, H.H. (1974a).
in man. In Handbook of Sensory Physiology,
vol. 6 (Vestibular System), part 2, H.H. Kornhuber (Ed.),
Springer-Verlag, Berlin, pp.193-232.
The vestibular system and the
Kornhuber, H.H. (1974b).
general motor system. Iu Handbook of. Sensory Physiology,
vol. 6 (Vestibular System), part 2, H.H. Kornhuber (Ed.),
Springer-Verlag, Berlin, pp.581-620.
The Organization of Voluntary Movement:
Kots, Y.M. (1976).
Neurophysiological Mechanisms, NASA Technical Translation,
NASA-TT-F-16871, original in Russian, "Nauka" Press, Moscow,
1975.
A signal-to-noise
Kreifeldt, J.G. and Yao, S. (1974).
investigation of nonlinear electromyographic processors.
IEEE Trans. Biomed. Eng. BME-21:298-308.
Muscle responses
Lacour, M., Xerri, C., Hugon, M. (1978).
The role of
and mono-synaptic reflexes in falling monkey.
the vestibular system. J. Physiol.(Paris) 74:427-438.
Laursen, A.M., Dyhre-Poulsen, P., Dj/rup, A., and Jahnsen,
Programmed pattern of muscular activity in
H. (1978).
monkeys landing from a leap. Acta Physiol. Scand. 102:492494.
Lestienne, F., Soechting, J., and Ber'thoz, A. (1977).
Postural readjustments induced by linear motion of visual
Exp. Brain Res. 28:363-384.
scenes.
Visual
Levy, D.L., Proctor, L.R., and Holzman, P.S. (1977).
Otolaryngol.
Acta
interference on vestibular response.
103:287-291.
Cerebral organization and the conscious
MacKay, D.M. (1966).
control of action. In Brain and Conscious Experience,
J.C. Eccles (Ed.), Springer-Verlag, New York, pp.422-445.
FuncMaeda, M., Magherini, P.C., and Precht, W. (1977).
tional organization of vestibular and visual inputs to neck
J. Neurophysiol.
and forelimb motoneurones in the frog.
4:225-243.
-
271 -
Marini, G., Provini, L., and Rosina, A. (1975).
Macular input
to the cerebellar nodulus. Brain Res. 9a:367-371.
Marsden, C.D., Merton, P.A., and Morton, H.B. (1977).
sensory mechanisms of servo action in human muscle.
J. Physiol. (London) 265:521-535.
The
Matthews, B. and Whiteside, T.C.D. (1960).
Tendon reflexes
in free fall. Proc. Roy. Soc. Series B 153:195-204.
McKechnie, R.E. and Cousins, S.J. (-1976).
A parallelogram
chain mechanism for measuring human joint motion. Presented
at the Design Engineering Technical Conference of The American Society of Mechanical Engineers, Montreal, Quebec,
25 June 1976.
McNemar, Q. (1969).
New York.
Psychological Statistics, Wiley,
Melvill Jones, G. and Watt, D.G.D. (1971a).
Observations on
the control of stepping and hopping movements in man.
J. Physiol. (London) 219:709-727.
Melvill Jones, G. and Watt, D.G.D. (1971b).
Muscular control
of landing from unexpected falls in man. J. Physiol.
(London) 219:729-737.
Melvill Jones, G. and Young, L.R. (1978).
Subjective detection of vertical acceleration: a velocity-dependent response?
Acta Otolaryngol. 85:45-53.
Modianos, D.T. and Pfaff, D.W. (1976).
Brain stem and
cerebellar lesions in female rats. I. Tests of posture and
movement. Brain Res. 106:31-46.
Morin, C. and Pierrot-Deseilligny, E. (1977).
Role of Ia
afferents in the soleus motoneurones inhibition during a
tibialis anterior voluntary contraction in man. Exp. Brain
Res. 21:509-522.
Nashner, L.M. (1970).
Sensory feedback in human posture
control. Sc.D. thesis, MIT.
Nashner, L.M. (1971).
A model describing vestibular detection of body sway motion. Acta Otolaryngol. 72:429-436.
Nashner, L.M. (1973).
Vestibular and reflex control of
normal standing.
In Control of Posture and Locomotion,
R.B. Stein et al. (Eds.), Plenum Press, New York, pp.291-308.
Nashner, L.M. (1976).
Adapting reflexes controlling the
human posture. Exp. Brain Res. 26:59-72.
-
272 -
Fixed patterns of rapid postural
Nashner, L.M. (1977).
Exp. Brain Res.
responses among leg muscles during stance.
-30:13-24.
Visual contribution to
Nashner, L. and Berthoz, A. (1978).
Brain Res.
rapid motor responses during postural control.
15 0: 403-407.
Influence of head
Nashner, L.M. and Wolfson, P. (1974).
latency postural
on
short
cues
proprioceptive
position and
reflexes evoked by galvanic stimulation of the human labyBrain Res. 67:255-268.
rinth.
Anatomical aspects of the funcNyberg-Hansen, R. (1975).
tional organization of the vestibulospinal pathways.
In The Vestibular System, R.F. Naunton, (Ed.), Academic
Press, New York, pp.71-96.
Oscarsson, 0.
channels from
physiological
M.D. Yahr and
pp.93-117.
(1967).
Functional significance of information
In Neurothe spinal cord to the cerebellum.
Basis of Normal and Abnormal Motor Activities.
D.P. Purpura (Eds.), Raven Press, New York,
Functional organization of spinocereOscarsson, 0. (1970).
bellar paths. Iti Handbook of Sensory Physiology, vol. 2
Springer-Verlag,
(Somatosensory System), A. Iggo (Ed.),
Berlin, pp.121-127.
Measurement of vestibulo-spinal
Peitersen, E. (1974).
of Sensory Physiology, vol. 6
In
Handbook
responses in man.
(Vestibular System), part 2, H.H. Kornhuber (Ed.), SpringerVerlag, Berlin, pp.267-280.
Properties of projecPeterson, B.W., and Abzug, C. (1975).
tions from vestibular nuclei to medial reticular formation
in the cat. J. Neurophysiol. 38:1421-1435.
A new long spinal
Peterson, B.W. and Coulter, J.D. (1977).
projection from the vestibular nuclei in the cat. Brain Res.
122:351-356.
Peterson, B.W., Maunz, R.A., and Fukushima, K. (1978).
Properties of a new vestibulospinal projection, the caudal
vestibulospinal tract. Exp. Brain Res. 32:287-292.
Electrical safety in the hospital.
Pfeiffer, E.A. (1977).
Technical report published by Scientific Enterprises, Inc.
Cerebello-vestibular interactions.
Pompeiano, 0. (1974).
In Handbook of Sensory Physiology, vol. 6 (Vestibular
System), part 1, H.H. Kornhuber (Ed.), Springer-Verlag,
Berlin, pp.414-476.
- 273
-
Pompeiano, 0. (1975).
Vestibulo-spinal relationships.
In The Vestibular System, R.F. Naunton, (Ed.), Academic
Press, New York, pp.147-180.
Precht, W. (1974a).
Characteristics of vestibular neurones
after acute and chronic labyrinthine destruction.
In Handbook of Sensory Physiology, vol. 6 (Vestibular System),
part 2, H.H. Kornhuber (Ed.), Springer-Verlag, Berlin,
pp.451-462.
Precht, W. (1974b).
The physiology of the vestibular nuclei.
In Handbook of Sensory Physiology, vol. 6 (Vestibular
System), part 1, H.H. Kornhuber (Ed.), Springer-Verlag,
Berlin, pp.353-416.
Prochazka, A., Schofield, P., Westerman, R.A., and Ziccone,
S.P. (1977).
Reflexes in cat ankle muscles after landing
from falls. J. Physiol. (London) 272:705-719.
Putkonen, P.T.S., Courjon, J.H., and Jeannerod, M. (1977).
Compensation of postural effects of hemilabyrinthectomy in
the cat.
A sensory substitution process?
Exp. Brain Res.
28:249-257.
Quandieu, P., Cailler, B., Pellieux, L., and Borredon, P.
(1977).
Determination of the biomechanical parameters of
man's vertebro-muscular axis.
Proceedings of the European
Symposium on Life Sciences Research in Space.
Cologne-Porz,
Germany, 24-26 May 1977.
(ESA SP-103:335-341).
Reschke, M.F., Anderson, D.J., Moore, M.J., and Homick, J.L.
(1976).
Vestibular-spinal responses as a function of zero-g.
Presented at conference of the Society for Neuroscience,
9 Nov 1976.
Roberts, T.D.M. (1967).
Labyrinthine control of postural
muscles. In The Role of the Vestibular Organs in Space
Exploration, NASA SP-152, Washington, pp. 160-180.
Roberts, T.D.M. (1978).
Neurophysiology of Postural Mechanisms, Butterworths, London.
Roberts, T.D.M. (1973)
Reflex balance.
Nature 244:156-158.
Rossignol, S. and Melvill Jones, G. (1976).
Audio-spinal
influence in man studied by the H-reflex and its
possible
role on rhythmic movements synchronized to sound.
Electroenceph. Clin. Neurophysiol. 41:83-92.
Rubin, A.M., Liedgren, S.R.C., Odkvist, L.M., Milne, A.C.,
and Frederickson, J.M. (1977).
Labyrinthine and somatosensory convergence upon vestibulospinal neurons. Acta Otolaryngol. 86:251-259.
- 274 -
Schaefer, H.P. (1974).
Compensation of vestibular lesions.
In Handbook of Sensory Physiology, vol. 6 (Vestibular
System), part 2, H.H. Kornhuber (Ed.), Springer-Verlag,
Berlin, pp.463-490.
Schomburg, E.D., Meinck, H.M., Haustein, J., and Roesler, J.
(1978).
Functional organization of the spinal reflex pathways from forelimb afferents to hindlimb motoneurones in the
cat. Brain Res. 1_3.:21-33.
Responses of cat vestibular neurons to
Schor, R.H. (1974).
sinusoidal roll tilt. Exp. Brain Res. 20:347-362.
Neurophysiology of
Shik, M.L. and Orlovsky, G.N. (1976).
locomotor automatism. Physiol. Rev. 5-6:465-501.
Visual climbing fibre
Simpson, J.I. and Alley, K.E. (1974).
input to rabbit vestibulo-cerebellum: a source of directionspecific information. Brain Res. 82:302-308.
Smith, J.L., Edgerton, V.R., Betts, B., and Collatos, T.C.
(1977).
EMG of slow and fast ankle extensors of cat during
posture, locomotion and jumping.
J. Neurophysiol. 140:503513.
Soechting, J.F., Anderson, J.H., and Berthoz, A. (1977).
Dynamic relations between natural vestibular inputs and
activity of forelimb extensor muscles in the decerebrate
III. Motor output during rotations in the vertical
cat.
Brain Res. 120:35-47.
plane.
RelaSoechting, J.F., Burton, J.E., and Onada, N. (1978).
tionships between sensory input , motor output and unit
activity in interpositus and red nuclei during intentional
movement.
Brain Res. 152:65-79.
Sperry, R.W. (1950).
Neural basis of the spontaneous optokinetic response produced by visual inversion. J. Comp.
Physiol. Psychol. 413:482-489.
Peripheral control of movement.
Stein, R.B. (1974).
Physiol. Rev. 51:215-243.
Reflex responses of human
Stein, R.B. and Bawa, P. (1976).
J. Neurophysiol.
soleus muscle to small perturbations.
3:1105-1116.
Stott, F.D. and Weller, C.
using integrated circuits.
(1976).
Biomedical amplifiers
Med. Biol. Eng. 14:684-687.
Minimizing electrode
Tam, H.W. and Webster, J.G. (1977).
IEEE
Trans. Biomed. Eng.
abrasion.
by
skin
motion artifact
BME-24:134-139.
- 275 -
Takahashi, M., Igarishi, M., and Reschke, M. F. (1978).
Directional conflict between vestibular and visual inputs
in the squirrel monkey. Acta Otolaryngol. 85:253-261.
Teuber, H.-L. (1966).
Alterations~of perception after brain
injury. In Brain and Conscious Experience, J.C. Eccles
(Ed.), Springer-Verlag, New York, pp.182-216.
Thach, W.T. (1970a).
Discharge of cerebellar neurons related
to two maintained postures and two prompt movements.
I. Nuclear cell output. J. Neurophysiol. 33:527-536.
Thach, W.T. (1970b).
Discharge of cerebellar neurons related
to two maintained postures and two prompt movements.
II. Purkinje cell output and input.
J. Neurophysiol.
33:537-547.
Thach, W.T. (1978).
Correlation of neural discharge with
pattern and force of muscular activity, joint position, and
direction of intended next movement in motor cortex and
cerebellum. J. Neurophysiol. 41:654-676.
Thoden, U., Dichgans, J., and Savidis, T. (1977). Directionspecific optokinetic modulation of monosynaptic hind limb
Exp. Brain Res. 30:155-160.
reflexes in cats.
Tokita, T., Taguchi, T., and Matuoka, T. (1972).
A study
on labrynthine ataxia with special reference to proprioceptive reflexes. Acta Otolaryngol. L4:104-112.
Uemura, T. and Cohen, B. (1973).
Effects of vestibular
nuclei lesions on vestibulo-ocular reflexes and posture in
monkeys.
Acta Otolaryngol. SuPpl.
15:1-71.
Vallbo, A.B. (1970).
Slowly adapting muscle receptors in
man. Acta Physiol. Scand. 78:315-333.
Vidal, P.P., Lacour, M., and Berthoz, A. (1979).
Contribution of vision to muscle responses in monkey during free
fall: visual stabilization decreases vestibular-dependent
responses.
Exp. Brain Res. 31:241-252.
von Holst, E. (1954).
Relations between the central nervous
system and the peripheral organs.
Brit. J. Anim. Behav.
2:89-94.
von Holst, E. and Mittelstaedt, H. (1950).
Das Reafferenzprinzip (Wechselwirkungen zwischen Zentralnervensystem und
Peripherie).
Naturwiss. 31:464-476.
Waespe, W. and Henn, V. (1977).
Neuronal activity in the
vestibular nuclei of the alert monkey during vestibular and
optokinetic stimulation.
Exp. Brain Res. 21:523-538.
-
276 -
Waespe, W. and Henn, V. (1978).
Conflicting visual-vestibular stimulation and vestibular nucleus activity in alert
monkeys. Exp. Brain Res. 33:203-211.
Waespe, B., Waespe, W., and Henn, V. (1980).
estimation during visual-vestibular conflict.
publication in Arch. Psychiat.
Velocity
Submitted for
Watt, D.G.D. (1976).
Responses of cats to sudden falls:
an otolith-originating reflex assisting landing.
J. Neurophysiol. 319:257-265.
Webster, J.G. (1977).
Interference and motion artifact in
biopotentials.
IEEE 1977 Region 6 Conference Record,
pp.53- 6 4.
Wilson, V.J. and Melvill Jones, G. (1979).
tibular Physiology, Plenum Press, New York.
Mammalian Ves-
Wilson, V.J. and Peterson, B.W. (1978).
Peripheral and central substrates of vestibulospinal reflexes.
Physiol. Rev.
L_:80-105.
Wilson, V.J. and Yoshida, M. (1969).
Comparison of effects
of stimulation of Deiter's nucleus and medial longitudinal
fasciculus on neck, forelimb, and hindlimb motoneurones.
J. Neurophysiol. 32:743-758.
Wist, E.R., Diener, H.C., Dichgans, J., and Brandt, T.
(1975).
Perceived distance and the perceived speed of self
motion: linear vs. angular velocity? Perception and Psychophysics 17:549-554.
Yamada, H. (1970).
Strength of Bioloaical Materials,
Williams and Wilkins, Baltimore, pp. 23, 75-77.
Young, L.R. (1974). Role of the vestibular system in posture
and movement. In Medical Physiology, vol.1, V.B. Mountcastle
(Ed.), Mosby, St. Louis, pp.704-721.
Young, L.R., Oman, C.M., and Dichgans, J.M. (1975). Influence
of head orientation on visually induced pitch and roll sensation. Aviat. Space Environ. Med. 46:264-268.
Zacharias, G.L. and Young, L.R. (1977).
Manual control of
yaw motion with combined visual and vestibular cues.
Presented at Thirteenth Annual Conference on Manual Control,
MIT, Cambridge, Mass., 15-17 June 1977.