A Square Wave Electromagnetic Flowmeter
for Application to Intact Blood Vessels
By ADAM
B. DENISON, J R . , M.D., MERRILL P. SPENCER, M.D.,
HAROLD D. GREEN, M.D.
AND
A new type electromagnetic flowmeter which utilizes an alternating square wave magnetic field
applied to the unopened blood vessel is described. It attains stability of baseline and sensitivity
not possible with older D.C. and A.C. meters and therefore puts measurement of blood flow through
the unopened vessel on a practical basis.
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
T
amplitude and polarity of the voltage produced is
proportional to the velocity and direction of flow.
Unfortunately, recording presents a problem since
the flow voltages are quite small and easily obscured
by electrode potentials. Amplification of the signal,
which is necessary, is troublesome for direct currents.
These difficulties can be minimized by careful design
but the basic problem remains.
An AC recording system using a magnet energized
with a sinusoidal alternating current avoids the
polarization difficulties mentioned. This AC may be
of any convenient frequency, though 60 cycles per
second is popular because it is readily obtainable.
With the AC magnet, the flow voltage is also AC
since it depends upon the strength of the magnet.
Therefore, the electrode potentials difficulty may be
obviated by using capacitor-coupled amplifier
circuits. However, another difficulty arises due to
the AC magnetic field itself. The AC voltage induced in the electrode leads passing through the
magnetic field is much lai-ger than the flow voltage.
Though it is 90° out of phase with the flow voltage
and can largely be balanced out, residual variations
still obscure the true flow voltage. Satisfactory
meters of this type have been made6-* but the
induced electrode voltage remains as a limiting
factor in the range of applicability, especially for
recording small flows. We have, in the past9, used
the AC system for recording flows in cannulated
vessels. With a magnet supply frequency of 200
cycles per second and with the magnet-electrode
assembly cast in a rigid block of plastic, it was
possible to obtain satisfactory stability to record
flows in the range of 5 to 150 ml per minute.
Attempts to adapt this system to the unopened
vessel were not successful to a practical degree,
however, because the alternating magnetic field
induces electrical eddy currents not only in the
blood but also in the wall of the artery and in
surrounding plasma or saline; this gives rise to large
electrode voltages. In the rigidly controlled eannulaelectrode assembly, these are steady enough to be
balanced out; but changes in the conductivity,
thickness or orientation of the artery in the intact
HIS paper deals with a new application
of the principle of electromagnetic
blood flow recording which avoids many
of the troubles inherent in previous systems2- ••4. The reader is referred to numerous
previous articles for details of the background
theory not covered here6 ' ' ' 7 - 8 . Briefly, the
electromagnetic flowmeter depends for its
operation upon the fact that if a conductor of
electricity is moved through a magnetic field
so as to cut the lines of force, a voltage will
be generated which is proportional to the velocity of the motion, the strength of the magnet,
and the length of the moving conductor. In
the case of blood flowing through a blood vessel, the length of the conductor is the diameter
of that portion of the blood stream at right
angles to the magnet field and to the axis of
flow. The voltage can be led off by electrodes
at either end of the conductor, i.e., the surface
of the vessel in this case.
Electromagnetic flowmeters have heretofore
been of two genenil ty|3es, according to the kind of
magnet used: "DC" or "AC". The DC system uses
a permanent magnet or an electromagnet energized
by direct current. This type illustrates the principle
of electromagnetic flow recording quite nicely, since
From the Department of Physiology and Pharmacology of the Bowman Gray School of Medicine of
Wake Forest College, Winston-Sulem, N. C.
Received for Publication : August 5, 1954.
This meter was demonstrated at the 1954 meeting
of the Federation of American Societies for Experimontal Biology1.
This investigation was supported by research
grants H-4S7 and H-1094 from the National Heart
Institute, of the National Institute of Health, Public
Health Service.
39
Circulaiion Re*tartht Volume III. January1955
ry 1955
40
ELECTROMAGNETIC FLOWMETER
FUNCTIONAL BLOCK DIAGRAM OF THE SQUARE-WAVE
ELECTROMAGNETIC FLOWMETER
Electrode
Mogntt
Asstmbly
30 Cycle
Square Wave
Oscillator
Amplifier
Amplifier
ond
Goin Control
Electrode
Bolonce
Power Supply
and Voltage
Regulators for
entire unit
Blanking Circuit
tor Mognet
Switching P u l m
Synchronizing
Circuit
30 cycle
Filter
(optional)
Amplifier and
Synchronous
Rectifier
D. C
Current
Amplifier
Recorder
Synchronizing
(and photing)
Circuit
for oscillator and
rectifier
120 v
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
Fia. 1. Block diagram of circuits for electromagnet flowmeter.
vessel-pickup unit change the net induced electrode
voltage, upset the balance and often result in large
zero drift; this difficulty is inherent in the AC
magnetic system.
It was felt that the electromagnetic flowmeter would not be practical for routine measurement of flow in intact vessels until a system
could be developed which had the inherent
characteristic of being immune to effects not
only of electrical eddy currents but also of
electrode polarization voltages. This paper
describes a system which appears to approach
this ideal. The improved design described here
makes non-cannulating electromagnetic flow
recording practical as well as improving the
linearity and stability of flow recording through
cannula-electrode assemblies.
DESCRIPTION OF METER
The square wave electromagnetic flowmeter
system uses a magnet energized by direct current periodically reversed in polarity. The
flow voltage is, therefore, a square wave which
can be amplified by an AC amplifier, thus
minimizing polarization and amplifier drift
problems. Electrode voltage pulses induced by
the reversing of magnet polarity are of such
high frequency and short duration that they
can be filtered out or removed by a synchronized switching circuit leaving approximately
90% of the cycle representative of the flow
alone.
The amplifier output is converted to DC in
a synchronous discriminator circuit to drive
the galvanometer of a direct writing recorder.
The direction of flow determines the polarity
of the deflection, so backflow is recorded below
and forward flow above the zero flow line.
With the attainment of a square magnetic
field, variations in electrode contact resistance,
vessel wall thickness and pulsation affect
mainly the switching pulse and do not disturb significantly the flow potentials recorded
during the intervals of constant magnetic
field.
Several circuit arrangements are possible
for applying the square wave principle and it
appears that the particular experimental needs
determine to some extent which is best. The
meter described here uses a frequency of 30
cycles per second and normally is damped for
mean flow recording on the Esterline-Angus
recorder. For recording instantaneous arterial
blood velocities a higher frequency may be
used. Figures 1-5 diagram the circuit currently used for mean flow recording. The
filter may be used to reduce interference from
60 cycles AC, cardiac potentials or large fluctuating polarization voltages if the flow is
small but since it makes the calibration somewhat non-linear we prefer not to use it.
Figure 6 gives the details of the magnetelectrode assembly as built for application to
the unopened vessel. The small electromagnet
and electrodes are oriented as indicated and
molded in plastic*. The exact position and
* Ward's Bioplastic, 3000 Ridge Road,
Rochester 9, New York.
East
Magnit Supply
Sq. w. otc-omp
Synchronizing Cktt.
Juml, 1934
.01
• Sjnch
pin 5
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
Blonk
Synch
pin 7
Fin. 2. The upper portion of this circuit is a symmetrical 25-cycle multivibrator and flip Hop to
actuate the 6AG7's. The lower portion is a similar multivibrator and sj'nchronizing pulse amplifier
for subsequent circuits. Both are synchronized to 30 cycles sec. by pulses derived from the 60 cycle
line; the magnet oscillator is actuated by a slightly delayed pulse and the oscillators are interconnected to maintain the same relative phase.
Mo; 28, 1994
*• 1000 A
M - million J V
Copocitits undtr 3 art in pf
obon Sort In^i
Fiu. 'A. The dilTerential amplifier in the upper portion of this circuit amplifies the signal derived
from the electrodes; tho 6SL7-6H6 gate circuit interrupts the signal while the magnet polarity is
boing reversed; thus the switching spike is not recorded. The lower (blanking) portion is a singleshot multivibrator and flip flop to actuate tho gate circuit; it disables the amplifior for approximately
0.002 second at the beginning of each half of the square wave cycle.
41
42
ELECTROMAGNETIC FLOWMETER
From!
8H6 J
.1
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
Synch
pin 6
Synch
pin 3
Fio. 4. The upper circuits are another 6SL7-6H6 gate circuit followed by a conventional current
amplifier suitable for driving a commercial direct-writing recorder. The lower portion shows another
multivibrator-flip flop. This actuates the gate circuit so as to select half waves alternately from the
two sides of the push-pullflowsignal, thus accomplishing rectification and phase discrimination and
producing essentially a direct current output whose voltage and polarity reflect the velocity and
direction of the blood flow.
length of the electrodes is not critical so long
as the artery is between them at all times. This
may be assured by a thin plastic cap having a
ridge which fits into the groove and holds the
artery slightly away from the outer electrode.
Since the other electrode is in the bottom of
the groove the artery will always bear the
correct orientation to it.
Overall size of units now in use are approximately 23-~2 x 2 x 2 cms, tapered at the magnet
poles so that the groove is 8 to 10 mm in length.
A groove 5 mm. in depth and 2 mm. in width
accommodates the renal arteries of dogs up to
18 Kgm. For larger vessels the groove and
magnet should be increased in size so that the
flow is not restricted. When cannulation is
desired this assembly is modified so that the
groove is replaced by a plastic tube channel
in which the electrodes are oriented and through
which the blood is conducted from the vessel
cannula.
PERFORMANCE
Attaching the magnel-clcclrode assembly to the
vessel: The magnet-electrode assembly shown
in figure 6 is attached by freeing about three
centimeters of vessel from surrounding structures, placing the vessel in the indicated groove,
and securing the hinged cap. The electrodes
are then in close proximity to, or contacting,
the vessel wall. Actual contact is not essential
as long as all air is excluded from the spaces
surrounding the vessel. These spaces should
be filled with suitable electrolyte fluid such as
saline, plasma or blood. To accomplish this it
43
A. B. DENISON, JR., M. P. SPENCER AND H. D. GREEN
6AX5
120 »
J
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
Power Supply
June I, 1954
6V
DC
lor Ilrtt 3
GSC7'l in
omptilitr
+ 150
OK
+ IS0
amp
+ 130
Ml.
Fio. 5. Power supply for entire unit. Separate voltage regulator tubes are used for the difTerent
portions of the circuit as shown, to minimize interactions between them through the common power
supply.
is best to flood the region of the flowmeter
with such fluid. Further precaution in grounding nearby tissues at the site of application of
the assembly is desirable. The ground wire
should not be on a distant part of the body as
in that case potentials from the heart may be
disturbing.
Establishing the Zero Voltage-Zero Flow
Reference: The best baseline is one of zero net
electrode voltage and is established in three
stages while the flow is interrupted; first, by
centering the recorder to the desired position
without input; secondly, both sections of the
amplifier are centered to zero output, and
lastly the electrode voltages are balanced
until the recorder registers zero position. This
latter adjustment balances one electrode
against the other in the manner indicated in
the circuit diagram of figure 1, thus mini-
mizing residual disturbances caused by any
imperfections in the magnetic field wave form.
The accuracy of this adjustment can be checked
by switching the magnet off. If the recorder
continues on the same baseline the adjustment
is correct. When the blood is now allowed to
flow it will deflect the recorder pen according
to the direction and velocity of flow.
If the zero voltage reference was properly
adjusted the zero flow reference will agree with
it. During an experiment the zero flow reference may be checked by either of two methods.
The first of these is the simple expedient of
occluding the flow temporarily by constriction
of the vessel proximally or distally to the meter.
The agreement between the proximal and distal baseline illustrates the inconsequence of
electrode polarizing potentials caused by variations in contact with electrolyte or vessel wall
44
ELECTROMAGNETIC FLOWMETER
Plastic Casting
Slot for
vessel
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
3 2
Electrodes
o
Magnet
Coil • 4 0 0 0 turns
4 0 Enameled
Wire, Center Tapped.
Core- Laminated Silicon Steel
kj" x '/4" cross section
Fici. 6. Diagram of magnet-electrode assembly for
application to the unopened vessel. Numbers on
leads refer to corresponding numbers in detailed
circuit diagrams.
and proves that the position and shape of the
vessel segment tetween the electrodes is not
critical. Figure 7 illustrates this point in a
record taken during the measurement of renal
arterial blood flow where the zero flow following a strong epinephrine injection is seen to
agree with the zero reference obtained by
occlusion of the vessel upstream to the meter.
Stability of Zero Flow Reference: Drift of the
zero flow reference has been found to be no
greater than a deflection corresponding to 5
ml per minute over a two hour period for a
magnet-electrode assembly designed for an
artery carrying 100 to 150 ml of blood per
minute. In actual use occasional redetermination and adjustment of the zero reference further minimizes even this error.
Calibration Procedure: Calibration of the
flowmeter may best be accomplished by applying the unit to a convenient vessel in the animal, such as the carotid artery, which is cannulated distally for measurement of outflow
into a graduated cylinder. This method can
be used at the end of each experiment when
the animal is sacrificed. Another method
whereby the meter may be calibrated in situ
during the experiment is that of injecting
known volumes of blood, containing a strong
vasoconstrictor, downstream to the meter and
noting the backflow deflection of the recorder.
In vitro calibration can be performed conveniently by use of an excised vessel. Varying
quantities of saline or blood can be forced
through under pressure while the magnetelectrode assembly is immersed in saline. This
method of calibration is the one used in most
of the studies contained herein.
Linearity and Sensitivity: The calibration
pictured in figure 8 demonstrates the linearity
of this instrument for both forward and backward flow at two separate sensitivity settings.
The constancy of sensitivity from day to day
is also shown in this figure.
Due to differences in electrical conductivity,
the instrument is more sensitive to saline than
to blood but this factor does not affect the
linearity or stability. Since saline is cheaper
and easier to handle, most of the studies of
FIG. 7. Record of renal blood flow as measured with tho non-caunulating square wave electromagnetic flowmeter. (A) Control blood flow. (B) Epinephrine, 10 jigm injected intra-arterially. (C)
Occlusion of renal artery between the flowmeter and the aorta with a small clamp. (D) Power to magnet-electrode assembly turned off, demonstrating that zero reference is one of zero voltage.
A. B. DENISON, JR., M. P. SPENCER AND H. D. GREEN
40
y
DC 40 AC 60
DC 60 »C SOy"
30
45
tends to "ride" with the vessel and because
these disturbances are damped when mean
flow is recorded.
APPLICATIONS
20
• Mo?1 8 1954
S • Mar 191934
< l i U o ) 21,
10
•
0
-10
-ISO
-100 - 5 0
0
50
100
150
FLOW ml/minutt
I
200
I
250
I
300
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
FIG. 8. Calibration of Non-cannulating Electromagnetic Flowmeter O, • and O represent the calibration variation from day to day. X demonstrates
one of the many additional sensitivity settings available.
this report were, therefore, performed with
normal saline.
Effect of Variation in Vessel Size, Position
and Movement: The vessel size has no appreciable effect on the calibration of the instrument as shown by studies in which large and
small arteries and veins were perfused with
saline. The stability of sensitivity is dependent
primarily upon the fixed relationship between
the electrodes and the magnetic field so that
within the confines of the assembly groove,
variations in the size of the vessel or in its
position, relative to the electrodes, are not
critical. The instrument accurately integrates
pulsatile flow into mean flow. Calibration with
a pulsatile stream superimposes upon that of a
non pulsatile one. Pulsations of artery walls
have no net disturbing effect upon the mean
flow as it is measured and do not affect significantly the baseline of the zero reference.
This latter point is illustrated in figure 7 by
the downstream baseline established by epinephrine constriction. It is seen that the net
zero reference is not disturbed significantly in
spite of the remaimng pulsations.
If the vessel is pulled sharply through the
groove the baseline is momentarily disturbed
but quickly settles back to its original position.
This type of longitudinal movement of blood
vessels under experimental conditions is not
usually sufficient to be a disturbing factor
probably because the assembly is small and
Up to the present time this meter has been
used mainly on the renal artery of dogs10 but
also has been applied to the carotid, mesenteric and femoral arteries. The ease of application to a surgically remote vessel such as the
renal artery demonstrates one of its outstanding practical features. Theflowmeterdoes not
require that an anticoagulant be used, thus
avoiding hemorrhage as one of the factors in
the general deterioration of the animal. In
practice, the use of thisflowmeterin measuring
renal arterial inflow has been a gratifying
experience since the flow remains steady for
many hours.
Through the attainment of a zero flow
reference corresponding to zero electrode
voltage, this meter becomes applicable to long
term experiments in unanesthetized animals
where the small magnet-electrode assembly
may be implanted under sterile technique. The
incasement of the magnet-electrode assembly
and leads in plastic makes for easy sterilization, and the thermosetting plastic will withstand autoclaving without injury; this makes
possible the use of this meter in human surgery, particularly in peripheral vascular disorders and in human hemodynamic experimentation.
SUMMARY
An electromagnetic flowmeter principle is
described which is applicable to unopened
blood vessels and which avoids the sources of
instability characteristic of previous meters.
By imposing a square magnetic field on the
vessel and recording only the voltage generated
during the phases of constant magnetic flux
most spurious electrode potentials are eliminated. The square wave flow voltage suitably
amplified and rectified is proportional to the
volumetric rate and direction offlowand is of
a magnitude sufficient to be recorded on conventional direct writing recorders.
The following important advantages are
cited: (1) Zero drift is small. (2) The calibra-
46
ELECTROMAGNETIC FLOWMETER
tion for forward and backward flow is linear
and reproduceable. (3) Vessel size, position and
pulsation do not materially effect the operation of the meter Avhen mean flow is desired.
(4) The plastic magnet-electrode assembly for
attachment to the vessel is small, easily applied and can be sterilized. (5) Anticoagulant
is not necessary.
The meter accurately measures blood flow
through any large or medium sized vessel which
can be exposed surgically. Through modifications in construction of the magnet-electrode
unit many special needs of a variety of human
and experimental situations can be met.
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
REFERENCES
1
GREEN, H. D., DENISON, A. B., JB. AND SPENCEH,
M. P.: Vascular responses to adrenergic stimulation and blockade and demonstration of an
electromagnetic flowrneter for unopened blood
vessels. Fed. Proc. 13 (No 1. Part II) 643, 1954.
2
KOLIN, A.: An electromagnetic flowmeter,
principle of the method and its application to
blood flow measurements. Proc. Soc. Exper.
Biol. and Med., 36: 54, 1936.
1
—, An A.C. induction flow meter for measurement
of blood flow in intact blood vessels. Proc.
Soc. Exper. Biol. and Med., 46: 235, 1941.
MVETTERER, E.: New method of registering rate of
blood circulation in unopened vessels. Ztschr.
f. Biol. 98: 26, 1937.
5
KATZ, L. N. AND KOLIN, A.: The flow of blood in
the carotid artery of the dog under various
circumstances as determined with the electromagnetic flowmeter. Am. J. PhysioL, 122:
78S, 1938.
* KOLIN, A.: Improved apparatus and technique for
electromagnetic determination of blood flow.
Rev. Scient. Instruments, 23: 235, 1952.
7
JOCHIM, K. E.: Electromagnetic flow meter,
in Methods in Med. Research 1. Chicago,
The Year Book Publishers, 194S.
8
KOLIN, A.: A method for adjustment of the two
settings of an electromagneticflowmeterwithout
interruption of flow. Rev. Scient. Instruments,
24: 235, 1953.
' RICHABDSON, A. W., DENISON, A. B. AND GREEN,
H. D.: A newly modified electromagnetic
blood flowmeter capable of high fidelity flow
registration. Circulation 5: 430, 1952.
10
SPENCER, M. P., ROBERTS, G., AND GREEN, H. D.:
The direct renal vascular effects of epinephrine
and norepinephrine before and after adrenergic
blockade. Che Research 2: 537, 1954.
Effect of Blood Velocity on its
Spectral Transmission
Characteristics
In spectrophotometric studies of hemoglobin solutions the extinction of monochromatic light depends on the thickness of the transilluminated layer, the concentration of hemoglobin and the degree of its oxygenation. The presence of corpuscles
in whole blood causes extinction to deviate from Beer's law—that absorption of light
is directly proportional to the number of molecules that rays must penetrate. When
whole blood flows through tubes extinction is further affected by the velocity of flow.
The reason for the latter has recently been studied by Wever. On the basis of observations on whole bloodflowingthrough tubes and blood sedimenting by gravity, as well
1
as by other tests, he concludes that the effect of blood velocity on extinction is explained by changes in density of the axial stream.
He suggests that this effect of flow can be eliminated by placing in opposition two
photo cells which measure the passage of light in the red and infrared spectra respectively.
From R. Wever: Arch. ges. Physiol. 259, 97, 1954.
A Square Wave Electromagnetic Flowmeter for Application to Intact Blood Vessels
ADAM B. DENISON, JR., MERRILL P. SPENCER and HAROLD D. GREEN
Downloaded from http://circres.ahajournals.org/ by guest on October 2, 2016
Circ Res. 1955;3:39-46
doi: 10.1161/01.RES.3.1.39
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1955 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/3/1/39
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/