Quiz
Compare and contrast declarative (explicit)
vs. procedure (implicit or non-declarative)
memory. Address the following:
1. What is learned? Give examples.
2. What brain areas are involved in the
formation of memory?
3. Where is the memory stored?
PTP 512
Neuroscience in Physical Therapy
Motor Learning: Theories and Practical
Applications
Reading Assignments
Shumway-Cook: 21-39
Min H. Huang, PT, PhD, NCS
Outline for today’s lecture
•
•
•
•
•
•
•
•
•
•
•
Define motor learning and learning
Procedural learning
Declarative and associated learning
Adam’s theory
Schmidt’s theory
Ecological theory
Motor learning occurs in stages
Measuring learning outcomes
Transfer of learning
Feedback and giving augmented feedback
Practice conditions
DEFINING MOTOR LEARNING
Motor Learning
• Motor learning is the understanding of
acquisition and/or modification of
movement.
• As applied to patients, motor learning
involves the reacquisition of previously
learned movement skills that are lost
due to pathology or sensory, motor, or
cognitive impairments. This process is
often referred to as recovery of function.
PT Implications
• How can I best structure practice (therapy)
sessions?
• How often should my patient practice?
• Will the motor skill learned in one context
transfer to another?
• Will my patient be able to walk safely at
home/community after therapy?
• Should I simply the task?
• Should my patient practice weight shifting
in the // bars vs. walking in the gym (part
vs. whole practice)?
Learning vs. Motor Learning
• Learning is a relatively permanent
change in behavior due to practice, or
the process of acquiring knowledge
about the world.
• Motor learning: a set of processes
associated with practice leading to a
relatively permanent change in the
capacity for skilled actions.
Concepts of Motor Learning
• Learning is a process of acquiring the
capacity for skilled action
• Learning results from experience or
practice
• Learning cannot be measured or observed
directly; it is inferred from behavior
• Learning produces relatively permanent
changes in behavior; short term change is
not learning)
Motor Performance  Motor Learning
• Motor Performance is the temporary
change in motor behavior seen during a
practice session
– e. g. A patient learns how to shift more
body weight over the weaker leg at the
end of the therapy session. However, the
patient still bears more weight on the
unaffected leg at the next visit to PT.
Learning has not occurred.
Motor Performance  Motor Learning
• Performance may be influenced by many
other variables, e.g. fatigue, level of
learning/skills, anxiety, motivation, cues or
manual guidance given to the learner
• Motor Learning is a relatively permanent
change in motor behaviors that are
measured after a retention period and only
result from practice.
FORMS OF MOTOR LEARNING
Nondeclarative (Implicit) Learning:
Non-Associative Learning
• A single stimulus is given repeatedly and
the nervous system learns about the
characteristics of the stimulus
• Habituation
– ↓ response to the stimulus, e.g.
exercises to treat dizziness in patients
• Sensitization
– ↑ response to the stimulus, e.g. training
to enhance awareness of loss of balance
http://www.youtube.com/watch?v=voZXtTUdQ00
Nondeclarative (Implicit) Learning:
Associative Learning
• Classical Conditioning
– learn to predict relationships between
two stimuli
– e.g. before learning: verbal cues +
manual guidance  stand up; after
learning: verbal cue  stand up
– patients are more likely to learn if the
associations are relevant and meaningful
Nondeclarative (Implicit) Learning:
Associative Learning
• Operant Conditioning
– learn to associate a certain response,
from among many that we have, with a
consequence; trial and error learning
– e.g. relearn stability limits after ankle
sprain; verbal praise from PT
– behaviors that are beneficial and
rewarded tend to be repeated
Procedural Learning
• Does NOT require attention, awareness, or
other higher cognitive processes
• One automatically learns the rules for
moving, i.e. movement schema
• Learning requires repeating a movement
continuously under a variety of situations
• Patients with damage to cortex (e.g. TBI,
dementia, aphasia) can still  performance
Declarative (Explicit) Learning
• Require attention, awareness, and
reflection
• Results in knowledge or facts (e.g. objects,
places, events) that can be consciously
recalled and expressed in declarative
sentences, e.g. “1st I move to the edge of
chair. 2nd I lean forward and stand up”;
instruction from PT; mental rehearsal;
motor imagery
Declarative (Explicit) Learning
• Practice can transform declarative into
procedural or nondeclarative knowledge
– e.g. a patient first learns to stand up may
verbally repeat the instruction; after
repeated practice, the patient may be
able to stand up without instruction
• Processes of declarative learning: encoding
 consolidation storage retrieval
Think-Pair-Share
When helping a patient to relearn motor
skills, should the PT emphasize nondeclarative (implicit) or declarative explicit)
learning?
THEORIES OF MOTOR LEARNING
Adams Closed-Loop Theory
• In motor learning, sensory feedback from
the ongoing movements is compared with
the stored memory of the intended
movement
• Memory trace selects and initiates a
movement
• Perceptual trace, built-up over practice, is
the internal reference of correctness
Adams Closed-Loop Theory
• Clinical Implications
– Accuracy of a movement is proportional
to the strength of the perceptual trace
– Patient must practice the movement
repeatedly to ↑ the perceptual trace
• Limitations
– Cannot explain open loop movement or
novel movements
Schmidt Schema Theory
• Emphasizes open-loop control processes
and generalized motor program
• “Schema” is a generalized set of rules for
producing movements that can be applied
to a variety of contexts
• Equivalent to motor programming theory of
motor control
Schmidt Schema Theory
• Information stored in short-term memory
after a movement is produced
1. Initial movement conditions, e.g. body
position, weight of an object, step height
2. Parameters of a generalized motor
program
3. Outcome of the movement, in terms of
knowledge of results
4. Intrinsic sensory feedback of the
movement
Schmidt Schema Theory
• Information stored in short-term memory is
converted into two schemas
1.Recall schema selects a specific response
and contains rules for producing a
movement
2.Recognition schema evaluates the
response correctness and informs the
learner about the errors of a movement
Schmidt Schema Theory
• Clinical Implications
– Variability of practice↑ learning and
generalized motor program rules
– Novel movement can be made accurately
based on previously learned rules
• Limitations
– Vague; no consistent research finding in
support of variable practice
– Cannot account for one-trial learning (In
the absence of a schema)
Ecological Theory
• Learning involves the exploration the
perceptual and motor workspace
1. Identify critical perceptual variables, i.e.
regulatory cues
2. Explore the optimal or most efficient
movements for the task
3. Incorporate the relevant perceptual cues
and optimal movement strategies for a
specific task
Ecological Theory
• Clinical Implications
– Patients learn to identify relevant
perceptual cues that are important for
developing appropriate motor responses,
e.g. identify relevant perceptual cues for
reaching and lifting a heavy glass: weight,
size, or surface of the glass vs. its color?
Fitts and Posner Three Stage
Model: Cognitive stage
• Learner activities
– Learn what to do
– Learn about the task and goals
– Require high degree of attention
– Select among alternative strategies
– Performance may be more variable
– Fast improvement in performance
– Develop a motor program
Fitts and Posner Three Stage
Model: Associative Stage
• Learner activities
– Refine the skills
– Refine a particular movement strategy
– Performance is less variable and more
consistent
– Cognitive monitoring decreases
– Improve the organization of the motor
program
Fitts and Posner Three Stage
Model: Autonomous Stage
• Learner activities
– Become proficient, save energy
– Attention demands are greatly reduced
– Movements and sensory analysis begin to
become automatic
– Able to perform multiple tasks, scan the
environment
– Ability to detect own errors improves
Implications for PT
• Motor learning probably occurs in stages
• Activities of the patient are different in
the different stages
• Activities of the therapist should be
different in the different stages
Systems Three-Stage Model
• Learners initially restrict degrees of
freedom (DOF) and gradually release the
DOF as the task is learned and the skills
improve
• Novice Stage
– Simplify movement by constraining joints
and ↓DOF, e.g. muscles co-contraction
– Less energy efficient
Systems Three-Stage Model
• Advanced Stage
– Gradual release of additional DOF
– More adaptive to different contexts
• Expert Stage
– All DOF released
– Efficient and coordinated movements
– Exploit the mechanical and inertial
properties of the limbs and the
environment
Gentile’s Two Stage Model
• Early stage
– Understand the task goals, develop
movement strategies, recognize
regulatory features of the environment
• Late stage
– Refine the movement, consistent and
efficient performance
– Closed skills become fixation/consistent
– Opened skills become diversification/
adaptive
Think-Pair-Share
A patient is learning to use a walker.
What would the patient be able to do at
each stage of learning?
Cognitive stage (early)Association stage (late)Automatic stage (late)-
APPLICATION OF MOTOR LEARNING
THEORIES
How to Measure Learning?
• To separate the relatively permanent
effects of learning from the transient effect
of practice, learning can be measured using
retention or transfer designs.
1.Test the subject after a retention interval,
typically >= 24 hr
2.Choose the same task (retention test) or a
variation of the task (transfer test)(e.g.
different speed or lighting conditions for
walking)
Practice Level: How Much?
PRACTICE, PRACTICE, PRACTICE
• Animal Studies
– 9,600 retrievals over 4 week period
(Nudo, 1996)
– 7,000 trials of food catching in 5 weeks
(Pavlides, 1993)
• Humans?
– Dose-response log-linear relationship
– Are patients getting enough practice?
Feedback (FB)
• FB is all the sensory information that is
available as a results of a movement
• Types by mode of delivery
– Intrinsic (e.g. proprioception)
– Extrinsic (e.g. instruction from PT)
• Types of FB by information provided
– Knowledge of results (KR)
– Knowledge of performance (KP)
Knowledge of Performance (KP)
• Information about the movement patterns
• Usually intrinsic but can also be extrinsic
• Proprioception, Biofeedback, video
recording, verbal instruction (e.g. “Your
elbow was too low.”)
Knowledge of Results (KR)
• Information about the result or outcome of
the movement in terms of the goal
• Verbal instruction (e.g. “You were off the
target.”), proprioception (e.g. feeling loss of
balance during a fall)
Characteristics of Good Feedback
• Timing
– Allow some time to reflect between trials
• Summary FB
– Summary FB after a few trials works
better than after every trial
– Give more frequent summary feedback
(e.g. after every 5 trials) for complex
tasks than for simple tasks
Characteristics of Good Feedback
• Accuracy
– Positively reinforce correct performance
• Augmented (extrinsic) Feedback
– Video/visual of movement patterns alone
does not help; need to provide errorcorrecting cues as well
– AVOID VERBAL BOMBARDMENT
– Can be given concurrently or afterwards
Characteristics of Good Feedback
• Frequency and Fading Schedule
– More impaired patients may require
more frequent FB.
– Avoid giving FB every trial.
– Decrease the amount of FB given across
learning stages so the patients won’t
become dependent on FB.
Practice Conditions
• Massed vs. Distributed Practice Schedule
– Distributed in early stage (e.g. 20 min X 3
days) to avoid fatigue and massed in later
stage (e.g. 60 min in one day)
• Constant vs. Variable Practice
– Usually variable practice (walk at
different speeds) results in better
learning outcomes than constant practice
(walk at the same speed) in health adults
Practice Conditions
• Random vs. Blocked Practice
– Contextual Interference: practice that
makes the performance more difficulty
initially may result in more effective
learning in the long term
– Random practice (practice multiple tasks
in 15 min) results in better learning than
blocked practice (practice one task in 15
min) in healthy adults but not necessarily
patients
Practice Conditions
• Whole- vs. Part-Task Practice
– Task specificity says the best practice is
the task itself
– If utilizing a part technique, the part (e.g.
hip and knee flexion, extension) must be
a naturally occurring component of the
whole (e.g. walking)
Practice Conditions
• Transfer
– Amount of transfer is determined by the
similarity between the two tasks or the
two environments
– The more closely the practice
environment resemble those in the
performance environment, the better the
transfer
• Guided vs. Discovery Practice
Practice Conditions
• Mental Practice
– The same neural circuits producing the
movement are also active during mental
practice
– Can produce large positive effects on
performance of the task (Rawlings 1972)
– Physical + mental practice produces the
best learning outcome
Practice Conditions
• Action Observation
– Reorganization of the
primarily motor cortex in
patients with stroke after
observation of the
desired action + motor
training, but not
observation of irrelevant
movement + motor
training
– Implications for patients
with poor motor ability
Mirror neuron system:
Inferior parietal lobule
(IPL), Ventral premotor
cortex (PMv), inferior
frontal gyrus (IFG).
Garrison, 2010