|UFDC Home||myUFDC Home | Help|
GENERALIZATION OF REPETITIVE RHYTHMIC
CLAUDIA ANN RUTTER SENESAC
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Claudia Ann Rutter Senesac
This document is dedicated Emily Salles Senesac, Robert Edward Senesac and in
Memory of Ashley O'Mara Senesac and Robert Basil Rutter
I thank my family for their patience and understanding as I pursued a dream to
learn more about the body and mind. With their love and encouragement I complete this
journey and begin another.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ................................................................................................. iv
LIST OF TA BLES ........ .......................... ........ ...... ................ ............ vii
LIST OF FIGURES ..................... .......... ................. ................. viii
LIST OF OBJECTS ......... ........................... .......... .......... ........... ix
A B ST R A C T ................. .......................................................................................... x
1 IN TR OD U CTION ............................................... .. ......................... ..
Stroke and UE Rehabilitation ............. .... ........ ...... ...............
Theoretical Basis .......................... .............. .... ..... 5
Neuroplasticity and Use-dependent Plasticity ............................................................6
M otor L earning ...................... ............... ............. ... .............................. . 9
B ilateral Training................................................... 14
B iom echanics of R teaching ................................................................ ............... ..... 18
S u m m a ry .......................................................................................................1 9
S p ecific A im s ...............................................................2 1
Research Aim s and Hypotheses ........................................ ....... ............... 22
G e n e ra l a im 1 ......................................................................................... 2 2
Specific aim la and lb ...................................... ............................22
Primary hypotheses la and lb for spatial parameter............................. 22
Secondary hypotheses 2a and 2b for temporal parameters ........................23
2 M E T H O D S ......................................................... ................ 2 5
Experim mental D design .......................................... .. .. .... ........ .. ....... 25
S u b j e c ts .......................................................................................2 5
P ro c e d u re .......................................................................................................2 6
O utcom e M measures .............................................................................................. 29
Kinematic analysis of reaching ........................................29
N ovel Task #1(Sim ilar) ........................................ ................................30
N ovel Task #2(D issim ilar) ............................................................................ 31
Primary Spatial Dependent Variable.............. ...................................32
Secondary Temporal Dependent Variables ........................................................32
Posteriori of velocity profiles ............ ..................... ..................... 33
D ata A n aly sis ............................. ....................................................... ............... 3 4
3 R E S U L T S .............................................................................3 6
Data Analysis................................ .............. 36
Primary Spatial Dependent Variable..................................... ............... 36
H and path trajectory ..................... .. ..... ........... .. ...... .................... ..37
Secondary Temporal Dependent Variables ............. ....................................37
M ovem ent tim e .......................... .... .............. ................ ...........37
Time to peak velocity .............. ..... ........ ................... 38
P e a k v e lo c ity ................................................................................................. 3 8
A c c e le ra tio n ........................................................................................... 3 9
Posteriori of velocity profiles .. ........................... .....................39
D descriptive Individual D ata........................................ .................... ............... 39
4 DISCUSSION .................. .. .............. ..................49
Neuroscience Rationale for Repetitive Rhythmic Bilateral Training......................51
Factors Potentially Affecting the Study Results................................. ... ................ 53
Sum m ary ................ .......... ...... ... ....................................... 59
L im stations to the Study .......... .... .......................... ............. ....60
F utu re Stu dies ....................................................... 6 1
C o n c lu sio n s........................................................................................................... 6 2
LIST OF REFEREN CE S ........................................ ........................... ............... 64
BIOGRAPH ICAL SKETCH ...................................................... 72
LIST OF TABLES
2-1 D em graphic data ...................... ...................... ................... .. ...... 27
3-1 Means and standard deviations (sd) for all dependent variables ...........................41
3-2 Summary ANOVA model for the primary dependent variable ............................42
3-3 Summary of ANOVA model the secondary dependent variables............................42
3-4 Summary ANOVA model for posteriori analysis of velocity profiles...................43
3-5 Sum m ary of individual change ........................................ ......................... 44
LIST OF FIGURES
2-1 Illustration of testing conditions for similar spatial orientation novel task #1.........30
2-2 Illustration of testing conditions for dissimilar spatial orientation novel task #2 ....31
3-1 Significant main effect displayed for HPT for post-test.........................45
3-2 Significant main effect displayed for MT2 novel task #1................................ 46
3-3 Significant main effect displayed for PV on the post-test................... ............47
3-4 Significant main effect displayed for posteriori Peaks Metric...............................48
4-1 BATRA C invariant features ............... .................... .................... .. ........... 55
LIST OF OBJECTS
2-1 B A TR A C Inphase. ............................................. ................... .. .....29
2-2 B A T R A C A ntiphase ....................................................................... ..................29
3-2 Testing conditions for generalization Pre-test..................................................40
3-3 Testing conditions for generalization Post-test ............................................ 40
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
GENERALIZATION OF REPETITIVE RHYTHMIC BILATERAL TRAINING
Claudia Ann Rutter Senesac
Chair: Lorie Richards
Major Department: Rehabilitation Science
Background and purpose: Bilateral training (BT) is an alternative approach in
neurorehabilitation for individuals post stroke. Bilateral training activities may increase
the activity of the affected hemisphere and decrease the activity in the unaffected
hemisphere providing a balancing effect between hemispheric corticomotorneuron
excitability. One bilateral approach, repetitive rhythmic bilateral training, developed and
researched by Whitall, has shown improved motor function after intervention post stroke.
Yet, an important question is whether this type of practice will result in improvements in
untrained movements. The ability to perform related untrained motor tasks is
generalization. The purpose of this study is to determine if repetitive rhythmic bilateral
training will promote spatial generalization to a novel task.
Methods: Fourteen participants with hemiparesis completed the study. The
intervention used an arm training machine-BATRAC-consisting of two paddles mounted
in nearly frictionless tracks. The participants moved the handles back and forth in a
rhythmic manner for 5-minute blocks. Half of the blocks were in-phase; the other half of
the blocks were anti-phase. Practice sessions were 4 days/week, 2:25 hours/day, for 2
weeks, for a total of 18 hours of training. We measured movement time, time to peak
velocity, hand path trajectory, peak velocity, and acceleration using the Vicon motion
analysis system during 2 reaches to target tasks pre- and post- training. Each participant
gave informed consent according to University of Florida Institutional Review Board and
North Florida/South Georgia Subcommittee for Clinical Investigation requirements prior
Results: Improvements were found at post-test only for hand path trajectory and
peak velocity. They were equivalent across similar and dissimilar tasks. Movement time
2 was less for novel task #1 compared to novel task #2 but equivalent across pre- and
post- testing periods. No interaction effects were found.
Conclusion: Unlike Whitall, our kinematic results suggest that repetitive rhythmic
BT alone is not sufficient to change motor control, specifically generalization to similar,
but untrained tasks. However, the small and heterogeneous study sample precludes
definitive conclusions regarding the usefulness of this practice paradigm for promoting
motor skills post-stroke.
Stroke strikes 700,000 persons each year in the United States resulting in varying
degrees of permanent disablement.1 Many of the people affected by stroke will have a
residual upper extremity (UE) motor and sensory deficit that will influence their ability to
participate in life roles. These deficits typically involve decreased UE use and
coordinative control of the arm and hand for activities of daily living, gesturing, and
bilateral activities and will affect 78% of those individuals surviving stroke.2 Furthermore
the vast majority of those with severe UE paresis will not recover full function of their
arm and hand after 6-11 weeks of "traditional" therapy. 3 The effectiveness of current
rehabilitation approaches for restoration of UE function has not identified one
intervention as being superior to others in gaining function in the UE.3 4 Scientific
evidence now suggests that to enhance motor recovery post stroke, one of the critical
components in an intervention protocol is practice.5'6 Thus finding effective UE motor
rehabilitative interventions is an important goal.
Development of new, more effective rehabilitation techniques depends upon
understanding the neural, physical and behavioral expression of movement.7 Specifically,
an understanding of the CNS's ability to recover in the face of injury, and the extrinsic
factors that can influence that recovery, is essential for successful neurorehabilitation.
Key to motor recovery following stroke is the CNS' ability to learn or relearn motor
behaviors. This recovery can occur spontaneously, or more likely, will require practice
of the lost motor abilities to facilitate reorganization of the motor cortex.5'7'8
Re-training and practicing every motor behavior that the individual will be called
upon to use in everyday life is unrealistic. Motor control theory suggests that all
movement behaviors, or tasks, contain essential features that can be entrained with
practice.9 Motor learning theory further suggests that these features, once trained, can be
transferred to another task that requires the same features(s). This is called a
generalization effect.9Taking advantage of this effect could have a significant impact on
UE rehabilitation post-stroke.
Bilateral training is an emerging approach in neurorehabilitation for individuals
post-stroke. Bilateral movements form a tight phasic relationship organizing the behavior
to perform as a functional synergy.9-11 Animal and human research has supported the
notion that both hemispheres are active during bilateral activities. 1214 During the
acquisition phase of learning a bilateral skill, there is a functional coupling of motor areas
in both cerebral hemispheres.15 In persons post-stroke, bilateral activities have increased
activity of the damaged hemisphere and decreased activity in the undamaged to facilitate
a more balanced effect of between-hemisphere corticomotorneuron (CMN) excitability.16,
17 For example, repetitive bilateral arm training increased activation in the contralesional
cerebrum and ispsilesional cerebellum after 18 hours of training.18 The response of the
motor cortex to bilateral training with reorganization is encouraging. Bilateral training
may lay a foundation in individuals post stroke for engaging coordinative structures
allowing the execution of basic motions and movement even though the practice is not
real life task practice.
The basic motions provided in bilateral training may entrain both hemispheres and
provide the essential features necessary to generalize to similar tasks not specifically
trained in the intervention. Generalization of a task should be optimal when the neural
demands and conditions are similar.9 However, because functional tasks are not directly
trained, test of the generalization is important. The purpose of this study is to determine if
repetitive bilateral training will promote spatial generalization to a novel task.
The following literature review is composed of five main sections and will serve to
orient the reader to foundation principles underlying the purpose and hypotheses of this
project. The sections will include the following: 1) stroke and UE rehabilitation; 2)
theoretical basis; 3) neuroplasticity and use-dependent plasticity; 4) motor learning:
practice and generalization, and 5) bilateral training in stroke rehabilitation. First, the
traditional view of stroke and rehabilitation will be compared and contrasted with more
recent views based upon new scientific evidence. Second, conceptual frameworks for
studying UE recovery following stroke will be reviewed. In the third section, basic
elements of CNS neuroplasticity and training effects on CNS plasticity will be reviewed.
Motor learning principles including generalization, and practice and their relevance to
stroke rehabilitation will then be discussed in the fourth section. Finally, the use of
bilateral training incorporating key motor learning principles will be discussed as a
potential new therapeutic approach.
Stroke and UE Rehabilitation
Each year 700,000 persons will suffer a new or recurrent stroke in the United States
resulting in varying degrees of permanent disablement.19 Many of the people affected by
stroke will have a persistent upper extremity (UE) motor and sensory deficit that will
influence their ability to participate in activities of daily living and life roles. Motor and
sensory deficits typically involve decreased coordinative movement of the arm and hand
and UE use for activities encountered in a person's daily environment including self-help
skills, gesturing and bilateral activies.2 Furthermore, the vast majority of those with
severe UE paresis will not recover complete function of their arm and hand after 6-11
weeks of "traditional" therapy. 3 In fact, use of the UE is so important that subjective
measures of "well being" are directly related to perceived motor impairments of the arm
affecting quality of life. 19 Thus, developing effective UE motor rehabilitative
interventions is extraordinarily important, especially in light of current rehabilitative
approaches to UE treatment that lack clear consensus, and are conflicted and
Upper extremity interventions in the stroke population have historically focused on
treatment of single limb movements; treating the intact and affected arm separately.2024
Methods designed to restore motor skill in the affected UE are influenced by facilitation
models of motor recovery and emphasize handling or guidance to achieve more normal
movement patterns. Practice under these conditions improves performance, but improved
performance of a task does not necessarily lead to relatively permanent changes, which
characterize motor learning.7' 9 Performance during motor skill learning is a temporary
change in behavior that is observed during practice sessions but may not be retrieved at a
later time for execution.9 This may be because practice under traditional motor
rehabilitative approaches allows for few errors, and little problem solving (by the learner)
of the criteria inherent in the task.25 In addition many of these traditional therapy
approaches were based on the hierarchical-reflex theory, which has not held up under the
scrutiny of the current motor control and motor learning literature. Nonetheless, these
facilitory models continue to be used as traditional standard of care for rehabilitation.26-29
The hierarchical-reflex model suggests that motor learning in rehabilitation is a
stepwise sequence of motor recovery and motor development from lower levels to higher
levels of control.30 Treatment based on this theory of motor control often will focus on
the movements that are most automatic requiring sensory stimulation from the therapist,
progressing to more skilled voluntary tasks. This approach is often referred to as a
"traditional" approach utilizing facilitation from the therapist to accomplish goals. New
approaches to rehabilitation are beginning to surface based on the concepts of
neuroplasticity and motor learning theories.
Bernstein first proposed the systems theory in the early to mid 1900's although it
was not incorporated into rehabilitation until the early 1980's. 9,25, 31 Bernstein viewed
the nervous system as one of many contributors to movement execution but not the
controller of movement. Movements are seen as a result of an interaction among many
systems including internal and external environments, organized around behavioral goals
with distributed control.
Bernstein noted that there were many degrees of freedom (df) available to produce
a movement. Different df are characteristic of a task, environmental demands, and the
performer. These df need to be controlled for effective movement to be accomplished.
He purposed that the formation of synergies (groups of muscles and joints constrained
together) could control the multiple df problem. This model can describe how learning of
a new motor skill takes place. In the early stages of learning a new task the movement
may be simple. Movement at one joint may be allowed to vary with intermediate joints
held stiffly utilizing cocontraction of muscles to control the df. Once the movement is
learned muscle cocontraction is reduced and the movement becomes more fluid
indicating the ability of the central nervous system to use multiple resources to
accomplish a task. In stroke, these synergies are constrained in a pathological manner
with stereotypical movements observed with attempts to move.21' 22 Difficulty is noted in
the ability to control the multiple dfthat are available in the extremities and trunk. This
inability to form normal synergies of movement leads to compensation and decreased
fluidity of movement in persons post- stroke.24
Neuroplasticity and Use-dependent Plasticity
Brain infarction results in a semi-reversible set of pathophysiological events
including swelling of the affected area, impaired circulation and pyramidal cell injury or
death.8 Recovery from brain infarction involves plasticity-the ability of the central
nervous system to reorganize after brain injury. 32 Developing an understanding of the
post-ischemic plasticity and its effect on motor control and motor learning has become
the focus of current rehabilitative efforts.9 The CNS post-stroke begins a process of
spontaneous recovery which involves neurological reorganization. In contrast to
individuals with intact nervous systems, attempts to move after stroke result in decreased
activation in the affected motor cortex with increased activity in the non-affected
hemisphere.8 33 34 These findings suggest even though crossed motor pathways are
damaged, after stroke recruitment of preexisting uncrossed motor neural pathways may
be accessed.34 Ipsilateral motor unit activity can be induced when the ipsilateral dorsal
premotor cortex area is stimulated by TMS. This stimulation has demonstrated shorter
latencies when compared to contralateral stimulation of the premotor cortex when the
hand is moved in stroke patients.35 Even in individuals who have recovered from stroke,
there is an increased activation of the CMN pool in the undamaged hemisphere when
compared to persons with intact nervous systems performing a finger tapping
movement.33 36 However, recovery from stroke is associated with decreasing activation
in the contralesional hemisphere and increasing activation in the lesioned hemisphere; a
more normal balance of activation is seen.
Motor recovery post stroke is augmented by rehabilitation. Rehabilitation of
individuals post-stroke involves motor learning. Motor learning is characterized by a set
of processes that are associated with practice. These processes influence change in the
internal state of the central nervous system and become relatively permanent, capable of
being retrieved from long-term memory centers into working memory for motor
execution.5 7 9 In the case of the individual with stroke, rehabilitation is concerned with
the relearning of once familiar motor skills using new motor pathways. Coordinative
patterns of movement must be practiced to create these new motor pathways during
recovery for the execution of motor skills. The capability by which the brain modifies
structure and function in response to learning or brain damage is neuroplasticity.5 7
Several mechanisms of reorganization of cortical areas after stroke or brain injury
have been proposed. Unmasking is a term used to indicate decreased inhibition of pre-
existing excitatory synapses allowing for functionally inactive connections to become
active. Changes that are rapidly induced during spontaneous recovery after injury are
believed to be unmasking.5 Synaptogensis refers to growth of new neural connections and
is related to environment and practice.8 Long term potentiation (LTP) involves the
increasing sensitivity of synapses pre-synaptically through constant stimulation with a
resultant larger postsynaptic output. Long term potentiation and synaptogensis are
believed to occur over longer time periods, coming into play during intense practice.5
Sparing refers to the areas of the brain that were not damaged during the injury and may
be adjacent or interconnected to the damaged area. These areas of sparing have been
shown to play a role in reorganization of cortical maps.5' 37-39
Nudo et al. has demonstrated in a study with squirrel monkeys that the motor
cortical areas of the brain can reorganize after brain injury.32 Further investigation by this
group of researchers suggest that reorganization of cortical maps is dependent on
rehabilitation and practice.38 40-42 Use-dependent plasticity relies on activation of the
brain during periods of practice.57 Calautti and Baron reviewed neuroimaging studies of
individuals post-stroke and found reorganization of motor areas with enhanced activity in
existing neural networks. In this review, both motor training and pharmacological
interventions were found to induce this increased activity in the damaged hemisphere
associated with recovery of function and improved motor skills.43 Calautti and Baron
observed significant changes in neuromapping in individuals involved in intense practice
of specific tasks. For example, neuroplasticity associated with motor rehabilitation has
been documented with the treatment paradigm constraint-induced movement therapy
(CIMT).44-46 Cortical reorganization in motor output areas of the damaged hemisphere
and in areas adjacent to the damaged site have been demonstrated as a result of intense
practice.37 47 48 Jang et al. demonstrated cortical reorganization by fMRI after 4 weeks
(4 days/week, 40 minutes/day) of task oriented training (practicing of functional
tasks). This training consisted of six tasks to improve UE function in 4 individuals with
chronic stroke. Cortical reorganization was evident with changes in the activation of the
primary sensory motor cortex (decrease in activation of the unaffected hemisphere and
increase activation in the affected hemisphere).49 These studies indicate that practice is an
important component of the motor rehabilitative process as it facilitates neuroplasticity of
the cortical motor areas.
Motor learning and the underlying neuroplastic changes are dependent on
practice. 5,7 In the face of neuropathology, persons post-stroke must confront
relearning skills that once were part of their daily routine. Determining the conditions for
optimal learning in persons post-stroke requires an understanding of the different types of
practice available during a rehabilitation program. The type of practice schedule that is
selected during therapy has a strong effect on the process of motor learning effecting the
basic components of movement and building the specifics of coordination for activities.9
Certain principles of motor learning are well established in healthy adults but not well
understood in stroke.
Mass practice builds capacity (skillfulness, ability) by utilizing longer practice
periods and short rest periods between trials. However, this type of practice can lead to
fatigue resulting in detrimental results for actual motor learning, transfer (generalization),
and retention.9 Practicing under conditions of fatigue may affect the synergies that are
engaged during the learning of the task, and ultimately the ability to retrieve the
appropriate information for the execution of the task at a later time (retention and
transfer) is reduced. Distributed practice provides shortened practice sessions and equal
or longer rest periods than the actual task trials. This type of practice improves
performance without the complication of fatigue and has demonstrated positive effects on
motor learning as measured by transfer trials. 9
The sequence of practice is also an essential component to motor learning. Practice
that repeats one task for a set number of trials before moving on to the next task is
referred to as blockedpractice. This type of practice has low contextual interference
(learning within the context of only one task) as the person learns the criteria for the task
and performs several repetitions in the acquisition phase before moving on to another
skill. Blocked practice enhances performance but may be detrimental to retention or
permanent learning as there is a low demand on problem solving once the criteria of the
skill are understood. Blocked practice is used in therapy when the task is just being
introduced and the participant is becoming familiar with the criteria of the movement.
Random practice intermixes trials so that no task is repeated on two consecutive trials.
The order of presentation of trials is varied presenting a high degree of contextual
interference (learning one task in the context of other tasks). Random practice provides
different patterns of coordination with different underlying motor programs with a range
of solutions for motor tasks.50 This type of practice can be detrimental to performance
during the acquisition phase but beneficial to motor learning and retention with the
continual demand on retrieving the criteria of the task.51' 52
During rehabilitation, therapists attempt to help the individual with stroke build the
ability to produce coordinative movements. Because it is impossible during rehabilitation
to practice every motor task the person will encounter in daily activity, it is believed that
the basic coordination gained by practicing some tasks during rehabilitation will
generalize to unpracticed motor tasks the individual will encounter in his/her everyday
life. The concept of "generalization" or transfer allows for the execution of other, related
skills apart from the specific practiced task (new skill or new environment). 7 9 The
critical aspect of generalization of a motor task appears to be whether similar neural
processing requirements of the tasks are incorporated. The more closely linked the
conditions and demands of the new skill or new environment are to the practice
environment, the better the transfer.53' 54 This ability to retrieve information for retention
and generalization is directly linked to practice and the type of invariant features that
constitute the motor skill.9
Choosing the right practice schedule and sequence are dependent on the stage of
learning that the individual is in and the classification of the motor skill that is to be
practiced. When a person who is neurologically intact is learning a new task he/she
begins by gaining an understanding of the rules and strategies inherent in performing the
task. Systems theory suggest that learning the invariant features of the task is
accomplished by engaging coordinative structures (muscles, joints, neural components,
arousal, and gravitational influences).7 31 The learning of a coordinative movement
involves components of the internal and external environment all of which contribute to
the pattern of coordination that emerges. Skilled movements require parameters that
make the task unique and different from other tasks. The unique features contain rules
that are particular to that task. Learning these rules and the invariant features of the task
is often referred to as the cognitive phase of learning.9 Each phase of learning allows for
the complexity of the task to increase and the motor skill to be refined until it is
In the early stages of recovery after stroke many people have difficulty initiating
any movement. This lack of movement, related to a decrease of the CMN pool, makes
activation and muscular recruitment a difficult task.55 Rehabilitation post stroke at this
level is concerned with gaining an ability to move, learning the basic interjoint
coordination and activation pattern of muscles, which gives feedback about movement
and how to recruit muscles for motor tasks. In persons post-stroke this stage of learning is
coupled with building the physiological capacity to move and is usually very cognitively
demanding, requiring high levels of concentration.
What do we know about practice and therapy intervention post-stroke? There are
few studies examining motor practice parameters post-stroke. Many of the studies that
are cited in the literature mention practice but fail to elaborate on the specifics of the
practice conditions using traditional therapy as the intervention.2729There are limited
studies on UE practice protocols in stroke that have shown positive results paying
attention to the specifics of practice. Some of these studies have demonstrated that some
of the same principles of motor learning and practice as established in healthy individuals
apply to motor learning post-stroke. Others have not. For example, Hanlon52 1996
studied 24 subjects with chronic hemiparesis to determine the effect of different practice
schedules for the acquisition and retention of a functional movement sequence for the
involved UE. Subjects were randomized into three groups: control, blocked, and random
practice groups. The movement sequence involved a serial task that was alternated with
trials on three other tasks in the random group. The movement sequence was practiced in
two blocks of five trials in the blocked group. A significant difference was found between
random and blocked practice groups with random practice being more effective for
retention over time in individuals post stroke.52 These results in the stroke population
follow the principles of motor learning, retention and transfer in healthy adults.9
In contrast, Cauraugh56 2003 compared blocked and random practice sequences
combined with active neuromuscular stimulation trials in subjects with stroke. The
movements practiced included: wrist/finger extension, elbow extension, and shoulder
abduction. The results indicated motor improvement in both groups without a difference
between the two practice sequences. This study did not support what we know about
contextual interference associated with random practice in healthy individuals.56 It is
difficult to compare these studies as they used different types of tasks. However, the
results of these studies illustrate how little is known about the effects of practice
protocols, type of task practiced, in combination with the level of recovery of the
individual participating in practice. The rules for practice in individuals post stroke are
unclear and the factors affecting the results of practice in this population have not been
established.35 Are the concepts of motor learning based on neurologically intact
individuals relevant when persons post-stroke have difficulty initiating movement and
use pathological synergies to accomplish the movements that they do execute?
Principles of generalization post-stroke have had even less science. First, let's
examine what is known about generalization in healthy individuals in relation to UE
movement tasks. Generalization is an ability to execute another motor task not
specifically practiced. The ability to dissociate a learned motor skill utilizing coordinative
structures and features of the practiced task that are similar to but not part of the practiced
task would be an example of generalization. In a study by Sainburg et al.57 hand
movement directions were reported to generalize for movements made up to 36 degrees
to either side of the trained direction in individuals without neurological deficits.
Generalization beyond the region of training has been documented successfully when a
tight coupling of angle of gaze (visual field) and the position of the hand and shoulder are
provided.58 In both studies, the nervous system demonstrated an ability to use sensory
information to recalibrate the internal model formed by the practiced task. This
recalibration allows for a limited amount of adaptation by the musculoskeletal system to
a novel task. This agrees with our knowledge that the best generalization occurs when
the neural processing requirements are similar to that of the practiced task. Which
parameters of the task that are critical are not clear. Understanding the parameters that
enhance generalization following stroke can contribute to the design of treatment
Acquiring interventions that would generalize to skills not specifically trained in
practice sessions (trials) would be advantageous therapeutically. Identifying particular
intervention protocols and pairing them with the stage of recovery or learning that the
individual may be in could enhance their rehabilitation. By building capacity early on in
recovery and layering more complex skills as individuals post-stroke gain an ability to
move could lessen their overall UE disablement. Introducing the right intervention at the
right time may enhance motor learning, retention and generalization.
A new approach to UE stroke rehabilitation; bilateral training is beginning to be
investigated systematically and demonstrating some positive results. 16,59-62 Protocols in
these studies have used functional and non-functional tasks practiced bilaterally with
similar temporal and spatial requirements.60 61 63 What do we know about the brain and
Researchers believe that bilateral training may be a good approach in stroke
rehabilitation based on what we know about the brain and bimanual coordination.
Bilateral movements form a tight phasic relationship causing them to perform as a
functional synergy.10' 11The establishment of such coordinative structures during bilateral
movement may serve as a template and entrain the paretic arm during the movement
phase with the uninvolved hemisphere providing a pattern of firing for the involved
hemisphere.64 Generalization of task performance from one arm to the other is not a new
concept. For example, reaching movements generalize from the dominant arm to the
non-dominant arm in healthy individuals.65 How does reaching with one arm improve
function in the contralateral arm? When learning the dynamics of a reaching task the
neural representation for the dominant arm in the contralateral hemisphere may engage
neural elements for both arms. To assess the dependence of generalization on callosal
inter-hemispheric communication, Criscimagna-Hemminger et al.65 further investigated
transfer of the dominant UE to the non-dominant UE in a person with a commissurotomy.
The results were similar with generalization from dominant to nondominant arm
(unaffected UE to affected UE).65 What is the relationship to individuals with stroke?
Conceivably bilateral training may have a similar effect on individuals post-stroke with
improved transfer from the uninvolved UE to the affected UE. Bilateral practice may
also be beneficial because both hemispheres are active during bilateral actions perhaps
activating uncrossed tracts.12-14, 34, 35, 66 Evidence from animal and human research
supports the notion that a temporal interaction between hemispheres occurs in the motor
Gerloff et al.15 reviewed the functional coupling of the motor areas of both cerebral
hemispheres during bilateral learning. Interhemispheric interaction is particularly
important during the acquisition phase of the skill.15 Bilateral training may be an
appropriate starting place for rehabilitation after stroke.
There are three studies using repetitive rhythmic bilateral training, which have
demonstrated some transfer effects to other tasks incorporating specific critical elements
of the original training task. In these studies training involved non-functional tasks
performed bilaterally with similar spatial and temporal parameters for each extremity
using "rhythmic" synchronized inphase and antiphase movements. However, many of the
bimanual tasks that are used in daily life require a different contribution from each
extremity. Thus, it is critical to show generalization of this type of training to other useful
coordinative patterns utilized in daily activities as our arms are not always performing
with similar patterns of movement. One of the missing components to our understanding
of this intervention in stroke is generalization of the bilateral training.
The Whitall, et al.61study investigated the hypothesis that bilateral upper extremity
training with auditory cueing of a metronome would improve motor function in persons
who had suffered from stroke. The intervention involved a custom designed arm-training
machine (BATRAC-bilateral arm training with rhythmic auditory cueing), which allowed
for elbow and shoulder flexion and extension coordinated to a metronome set at a self-
selected speed. This study concentrated on a proximal effector system involving
shoulder and elbow joints while limiting trunk forward lean during reaching by a chest
restraint. The design was a single group pilot study with 14 subjects consisting of 20-
minute training sessions, 3 times per week for a 6-week period of intervention with a total
of 18 sessions. Each session consisted of four 5-minute periods alternating inphase and
antiphase movements using the BATRAC interspersed with 10-minute rest periods for
distributed practice. Results indicated significant improvement in motor performance on
the Fugl Meyer (FM) upper extremity section, significant improvement in performance
time on the Wolf Motor Function Test (WMFT) and significant increase in daily use of
the affected extremity on the Maryland Arm Questionnaire for Stroke after six weeks of
training with sustained improvement at 8 weeks after training cessation. 61
In a follow-up study, fMRI demonstrated increased hemispheric activation during
paretic arm movement with changes in the contralesional cortex and ipsilesional
cerebellum after training utilizing the BATRAC.18 Cerebellar activity has been identified
as a principal region for the control of bimanual coordination.14 Although the numbers of
subjects (6) who demonstrated changes in the Luft et al.18 study are small, it is
encouraging data that supports repetitive bilateral training. Bilateral training as a potential
therapeutic intervention has been bolstered by evidence of reorganization of the motor
cortex in individual's using BATRAC post stroke.
Stinear and Byblow 16 had individuals with stroke perform active movement of
the unaffected wrist, which drove passive wrist flexion-extension of the affected UE
using a manipuladum at a self-paced rhythm. Focus was placed on the distal effector
system, the wrist joint. Nine subjects of a heterogeneous group of stroke participants
practiced for 60 minutes a day for a total of 4 weeks with a random assignment into
groups of synchronous and asynchronous practice. Five of the nine participants
demonstrated improvement in motricity scores as measured by the wrist, hand and
coordination components of the upper limb section of the FM Assessment of Motor
Function. Postintervention transcranial magnetic stimulation (TMS) revealed a decrease
in the unaffected cortical map volume in the subgroup of five patients that improved in
motricity. The subjects that demonstrated significant results were a mix of persons with
cortical and subcortical lesions, acute and chronic stroke, mild and severe disability, and
had a combination of synchronous and asynchronous training. The results of this study
suggest that bilateral training promotes a balancing of between-hemisphere corticomotor
Although these studies have suggested that motor learning has occurred after
bilateral training they have not delineated the parameters or limits of motor skill
generalization with this practice. Schmidt9 suggests that the more similar the neural
demands are during novel tasks the greater the transfer.54 Specifics of the basic neural
elements involved in repetitive bilateral training for individuals post-stroke have not been
assessed in a transfer test. There are only a couple of studies on bilateral coordination
that delineated parameters important for generalization of a novel task in individuals who
were neurologically intact.67'68
Little is known about the principles of generalization of bilateral coordination in
healthy individuals except the studies mentioned above and to date there is nothing in the
literature involving the stroke population. In neurologically intact individuals Temprado
and Swinnen demonstrated generalization of a bilateral coordination pattern to a novel
pattern when the spatial relative phase (RP) (a variable that characterizes the spatial
relationship between two limbs) of the transfer task was similar but not to a task with a
different spatial RP.67 Muscle synergies engaged during interlimb coordination tasks are
influenced by spatial orientation. The symmetry of movement may be an important
factor in improving coordination in bilateral tasks.68 Determining the relationship of
training task parameters (spatial, angle of gaze, joint angles) to generalization in
individuals post stroke has not been specifically investigated.
Biomechanics of Reaching
Reaching is a functional task that requires control of multiple joints through
space.69 Kinematic measures of reaching have been utilized in studies of generalization to
document change in the pattern of reaching.58' 67,70 Biomechanical evaluations are
capable of capturing interjoint coordination, and movement composition thus indicating
quality of the reaching pattern.69 71 72 Biomechanical measures assessing temporal
aspects of reaching include but are not limited to; movement time, time to peak velocity,
peak velocity which indicates symmetry of the reach, strategy for reaching, and
acceleration. Kinematic spatial parameters of reaching include hand path trajectory (how
straight is the path to the target) during the reach.71 The literature on the biomechanics of
reaching has documented that reaching post-stroke is slower, discontinuous with many
movement reversals (stops and starts during the movement to the target), and the
trajectories are curved to the target.69' 71-73 Reliability and validity of these biomechanical
measures in the stroke population are not established to date.71 However biomechanical
assessment of reaching may provide an understanding of motor control and assist in the
evaluation of new therapies.
Motor learning and neuroplasticity are dependent on practice. The appropriate
type, duration, intensity, and frequency of practice to enhance motor learning,
generalizability, and motor recovery have yet to be determined in the stroke population.
What parameters should be emphasized in rehabilitation during practice sessions to
maximize generalization? It is not clear if task specific training is important in laying a
foundation for coordinative movement in persons after stroke. Persons post stroke have
difficulty moving and must relearn coordinative patterns to execute motor tasks. Perhaps
the focus should be on engaging coordinative structures that might provide a general
motor template to build physiological capacity and complex movements. Establishing a
motor capacity to move post stroke may provide the framework for generalization of a
practiced task. Supportive evidence on bilateral training and generalization suggests that
engaging similar spatial and temporal synergies may have positive effects on motor
learning in persons post stroke.16, 18, 61, 63, 67, 68, 70
Utilizing Whitall's protocol for repetitive bilateral training: can a "general
framework" of coordinative synergies be created divorced from a particular skill or task
that would underlie the basics of motion and movement capabilities? Whitall's protocol
used an arm training machine (BATRAC) for repetitive bilateral training in one
orientation with inphase and out-phase movements. The repetitive movement of the UE's
in this protocol engages similar synergies as real life reaching tasks bilaterally
accomplished in the workspace directly in front of the person. Stinear and Byblow using
a distal effector system demonstrated a balancing effect between hemispheres of
corticomotor excitability. How functional is repetitive bilateral training and to what
degree if any will this type of training assist an individual with stroke to execute tasks
that were not specifically trained but similar?
Therapy interventions have focused on simulation of tasks that would be performed
in the home and community. Therapists have emphasized building a repertoire of
movement skills that incorporate components necessary for other unpracticed motor
skills. Practicing every task that will be encountered by an individual once they are
discharged from rehabilitation is impossible. Identifying intervention protocols that
generalize to tasks unpracticed in the rehabilitation arena is essential. Generalization of
learned motor skills would enhance a person's ability to participate in life roles at home
and in the community by increasing the number of conditions and solutions to a host of
motor problems. Bilateral training may lay a foundation in individuals post-stroke for
engaging coordinative structures allowing the execution of "basic" motions and
movement even though the practice is not variable or with real life tasks. Collecting
kinematic data for assessment of generalization of this training may help us to understand
which reach parameters might change after bilateral training. Repetitive bilateral training
is distributed blocked practice, which avoids fatigue but allows for the criteria of the skill
to be learned without building endurance. This combination of practice may build a
"physiological capacity" for movement in persons post-stroke. The entrainment of both
hemispheres during bilateral training provides a functional coupling of the motor cortexes
especially during the acquisition phase of learning a bilateral task.15 Yet, the ability of
such training to transfer to functional tasks awaits testing.
The aim of this study is to determine if repetitive rhythmic bilateral training (using
the BATRAC as outlined in the Whitall et al.61 study) will generalize to a novel task that
is performed with similar neural demands.
To test the hypotheses that repetitive bilateral training using blocked-distributed
practice will demonstrate spatial generalization to a novel task with similar neural
demands in joint angles, workspace, visual gaze angles, and muscle timing. The
following research questions will be addressed in a single group repeated measures
design employing a pre-test and post-test period. Outcome measures will be taken prior to
intervention and at the end (completion) of 2 weeks of proximal repetitive bilateral
Research Aims and Hypotheses
General aim 1
To determine spatial generalization to a novel task after proximal bilateral training
for the affected upper extremity.
Specific aim la and lb
la. To determine if proximal bilateral training generalizes to a novel task (#1) that
is similar in shoulder/elbow joint angles, constraints of muscles and joints (coordination),
visual gaze angles and a workspace identical to the training.
lb. A secondary novel task (#2) will be tested with different joint angles, visual
gaze angles and workspace from the training motion.
Primary hypotheses la and lb for spatial parameter
la. Generalization will occur for novel task (#1) when tested in the same workspace
with similar joint angles as practiced for the proximal bilateral training intervention. At
the end of week two of proximal bilateral training: kinematic data for hand path trajectory
to the target. will demonstrate generalization for novel task (#1). Data will be compared
to baseline data with improvement for the above kinematic outcome predicted. Hand
paths to the target in stroke are variable and lack continuity..69 71,72 Therefore, based on
the literature hand path trajectory will be straighter. lb. Generalization will not occur for
novel task (#2) that is dissimilar in joint angles and workspace to training. At the end of
week two of proximal bilateral training: kinematic data for hand path trajectory will not
generalize for novel task (2#). Data will be compared to baseline data with no
improvement for the above kinematic outcome predicted.
Secondary hypotheses 2a and 2b for temporal parameters
Temporal parameters are assessed separately in this study because neither the
training nor the testing tasks emphasized speed. Therefore these parameters may not
change. Individuals post-stroke do move slower when compared to healthy individuals
so it is possible that the speed may be different after intervention although it was not
2a. Improvement in movement time, time to peak velocity, peak velocity and
acceleration will occur for novel task (#1) when tested in the same workspace with
similar joint angles as practiced for the proximal bilateral training intervention. At the
end of week two of proximal bilateral training: kinematic data for movement time, time
to peak velocity, peak velocity and acceleration (the percent of the reach that is
acceleration) will change. Data will be compared to baseline data with improvement for
the above kinematic outcomes predicted. Individuals post-stroke move slower during
reaching and often demonstrate a skewed profile in reaching with a shorter relative
duration in the acceleration phase, peak velocity is often lower compared to healthy
individuals and absolute time to peak velocity is shorter.69' 71Therefore, based on the
literature movement time will decrease, time to peak velocity will increase, peak velocity
will be higher and the percentage of reach that is acceleration will approach 50% of the
2b. Improvement in movement time, time to peak velocity, peak velocity, and
acceleration will not occur for novel task (#2) that is dissimilar in joint angles and
workspace to training. At the end of week two of proximal bilateral training: kinematic
data for movement time, time to peak velocity, peak velocity, and acceleration (the
percentage of the reach that is acceleration) will not change for novel task #2. Data will
be compared to baseline data with no improvement for the above kinematic outcomes
predicted. Movement time, time to peak velocity, peak velocity, and the percentage of
reach that is acceleration will not improve.
This study employed a single group, repeated measures design that included a pre-
test baseline and post-testing at the completion of two weeks of proximal bilateral
Fifteen participants with hemiparesis and UE motor deficits were recruited from
the Brain Rehabilitation Research Center's stroke database at the North Florida/South
Georgia Veterans Health System. One subject dropped out of the study due to unrelated
medical reasons. This database consists of individuals with stroke who have been
recruited to participate in rehabilitation studies from the North Florida/South Georgia
VA, Shands Hospital at the University of Florida, Shands Rehabilitation Hospital, Shands
Hospital at Jacksonville, Brooks Rehabilitation Hospital and the Brooks Center for
Rehabilitation Studies in Jacksonville. Nine of the subjects were male and 5 were female
with a mean age of 64.4 (sd = 13.3) years and a mean of 5.5 (sd = 3.9) years post-stroke.
Five of participants had right-sided lesions and nine had left-sided lesions. Demographic
and clinical data for the subjects are summarized in Table 2-1. Whitall1, Luft,2 and
Stinear and Byblow3 all demonstrated treatment effects with sample sizes of 9-14
Inclusion criteria were: 1) single unilateral stroke at least 6 months prior, 2) no
active drug or alcohol abuse, 3) able to follow 2-step commands, 4) no history of a
clinical ischemic or hemorrhagic event affecting the other hemisphere, and no CT or MRI
evidence of more than a lacune or minor ischemic demyelination affecting the other
hemisphere, 5) no history of more than minor head trauma, subarachnoid hemorrhage,
dementia, learning disorder, drug or alcohol abuse, schizophrenia, serious medical illness,
or refractory depression, 6) some active movement in shoulder and elbow with palpable
extrinsic forearm finger muscle recruitment. Exclusion criteria: 1) no movement in UE
or no palpable muscle recruitment in extrinsic finger extensor muscles, 2) scores > 3 on
the Motor Activity Log, indicating a high level of UE function, 3) spasticity greater than
2 on the Modified Ashworth Scale.
Each participant gave informed consent according to University of Florida
Institutional Review Board and North Florida/South Georgia Subcommittee for Clinical
Investigation requirements prior to participation.
UE motor function for novel task (#1) and (#2) was tested at the beginning of the
baseline period prior to intervention and after two weeks of proximal bilateral training.
All participants performed a session of baseline testing immediately prior to starting the
intervention (see Outcome Measures section below). The two-week intervention period
was followed immediately by post-testing of UE generalization of training to a novel
task. As in Whitall et al.,1 training was provided for 18 hours. However believing that
intensity is important,46 these hours were provided in 8 sessions of 2.25 hours each
across 2 weeks for a total of 18 hours. Short term upper extremity practice has been
shown to be effective in improving upper extremity motor function in persons post-
Table 2-1. Subject demographic data
Subject Age Gender Years Side of Lesion site Fugl-Meyer*
(years) poststroke Lesion Pre
sd= 13.3 sd 3.9
L CVA L MCA,
L CVA L subcortical
R CVA R MCA
R CVA R MCA ischemic
R CVA R MCA posterior
R cortical infarct
L CVA L MCA posterior
R CVA R superior hyrus,
L CVA L frontal lobe
L CVA L subcortical
L CVA L MCA infarct,
deep white matter
lobe, cortex F/P
L CVA L infarct insula
L CVA L MCA infarct
*Based on Fugl-Meyer scale (maximum score) =66,
CVA= cerebrovascular accident
Proximal bilateral training: The proximal bilateral exercise was identical to that
performed in the study by Whitall and colleagues.1 In this paradigm, participants were
seated facing a table on which was placed the arm training machine-BATRAC consisting
of two paddles mounted in nearly frictionless tracks.' The handles of the device are
horizontally oriented and cylindrical in shape. The participants grab the handles of the
paddles (with the affected hand strapped on as needed) and move the handles back and
forth in a rhythmic manner for 5-minute blocks with 10-minute breaks between blocks to
minimize fatigue. There was a chest plate that prevented the participant from leaning
forward with trunk flexion when the handles were pushed away from the person's body.
This chest plate was set at a distance of six inches from the table and the participant was
asked to keep his/her trunk against this plate during the intervention periods. The distal
stop on the BATRAC was lined up to the metacarpalphangeal joint (MCP) when the
intact arm and fingers were extended directly in front of the body over the track. This
corresponded to 80% of the reach. When active range of motion was limited at the elbow
joint the distal stop on the BATRAC was set at the wrist joint initially and progressed to
the MCP joint the second week. For half of the blocks, the participants moved the
handles symmetrically (in-phase); while in the other half of the blocks the participants
moved the handles 180 out of phase. These trials were alternated and balanced across
subjects and sessions. Because movement of one paddle is independent of the other
paddle, participants had to coordinate the movements of both UE's in order to achieve the
correct temporal and spatial movement relationships. Participants were encouraged to
move the full range of the exerciser and were assisted as needed by the researchers.
Participants were asked to assume a comfortable self paced movement speed at the first
session, which was maintained throughout the daily training session with the use of
auditory cues provided by a metronome set at the self-selected frequency. The
metronome was set at the beginning of each day of training to the participant's
comfortable pace to prevent holding the individual back from making progress.
Object 2-1. BATRAC Inphase.
Object 2-2. BATRAC Antiphase
The primary hypothesis stated that in individuals with chronic stroke, bilateral
training would generalize to a novel task that was similar in neural demands to the
training but not to the dissimilar novel task. The dependent variables were divided into
primary and secondary based on differences in spatial and temporal parameters. Knowing
from the literature that movements in the upper extremity are affected by abnormal
synergies post-stroke, we believed that the intervention (synchronous and alternating
movements bilaterally) would effect the spatial parameter to a greater extent than
temporal parameters by breaking up the abnormal synergies through more normal
interlimb coupling.8' 9 Therefore, the primary goal was to assess spatial generalization
after bilateral training. The primary spatial dependent variable was HPT. The secondary
hypothesis stated that bilateral training would improve the temporal dependent variables
(MT1, MT2, TPV, PV and acceleration) after 2 weeks of intervention. The procedures
for testing pre- and post-intervention follow.
Kinematic analysis of reaching
Participants were seated on a bench with the hip and knee angle at 90 degrees and
the feet flat on the floor. Each participant was asked to position their buttocks and back
against a straight edge held behind them to assure the same start position on the bench
each testing period. The affected UE was positioned at rest, palm down, on a table placed
in front of the person at the same distance as the intervention table including the chest
plate distance of six inches. The UE was in neutral shoulder flexion/extension, rotation
and adducted. The elbow was flexed as in the start position for the arm-training machine.
Novel Task #1(Similar)
The start position on a table in front of them was identical to the start position used
for the training with the BATRAC however the arm-training machine was not used
during novel task #1 nor was there a chest restraint. End targets at approximately 80% of
reach were marked on the table at the same arm reach length (elbow extension in front of
the body as measured to the metacarpalphalangeal joint or wrist joint determined during
intervention) as in the arm-training machine (BATRAC). A reflective marker was placed
on the target so the vicon motion analysis system could pick up the end point.
Participants moved their arm and hand to the target and returned to the start position five
times (Figure 2-1). The participants were not asked to point to the target but to simple
reach to the target. To minimize fatigue, there was a 30 second rest break between
Figure 2-1. Illustration of testing conditions for similar spatial orientation novel task #1
Novel Task #2(Dissimilar)
The start position was on a table aligned with the paretic shoulder directly in front
of the subject's body at the same distance from the trunk as in the BATRAC during
intervention. The subjects were asked to move their arm and hand to a target on the table
that was aligned horizontally with the start position and with the non-paretic shoulder at
the near edge of the table and then return to the start position five times (Figure 2-2). A
reflective marker was used so that the vicon motion analysis system could pick up the
end point. There was no trunk restraint used during this testing condition. To minimize
fatigue, there was a 30 second rest break between reaches.
Figure 2-2. Illustration of testing conditions for dissimilar spatial orientation novel task
#2 (dissimilar task)
Kinematics of reaches were videotaped using a 3-D movement recording system (8
camera Vicon system). Retro-reflective markers were placed on C7 and T10 vertebrae, the
acromion process, clavicle, sternum, upper arm, lateral epicondyle of the elbow, medial
epicondyle of the affected UE, forearm, wrist condyles, dorsum of the hand, MCP joint of
the index finger, and the index fingertip of the affected UE. The data was collected at 100
Hz. All data was averaged using the middle 3 trials for each novel task.
Kinematics were analyzed for hand path trajectories (end point paths measuring
straightness of the hand path), movement time, time to peak velocity, peak velocity and
acceleration as in Cirstea, et al., Cunningham et al., and McCrea et al.9-12 Posteriori
analysis of the velocity profiles to assess movement smoothness changes were performed
as described by Rohrer et al.13
Primary Spatial Dependent Variable
1. To determine hand path trajectory (HPT), the ratio of the length of the actual path
traveled by the index finger in three-dimensional space to the length of an ideal straight
line joining the initial and final index finger positions was computed. If the participant
was unable to extend the index finger, the trajectory was measured from a marker on the
metacarpalphalangeal joint of the 2nd digit. The length index rather than the more usual
perpendicular distance between the trajectory better captures trajectories that may deviate
from the ideal straight line and may even intersect with that ideal line. Reach accuracy
was computed as the root mean squared error of the absolute distance between the final
endpoint position and the position of the target.
Secondary Temporal Dependent Variables
Temporal parameters in reaching post-stroke are significantly slower than healthy
adults.10-12 Therefore, although subjects were not asked to reach as quickly as possible
during the testing tasks, the predictions are based on the assumption that the training
intervention would increase their usual speed.
1. Movement time was the difference in time from movement onset to movement
offset. Movement time 1 (MT1) was defined as the onset of movement from the start
position to the touch of the target. A mark was made with tape to delineate the start
position and target area based on the intervention position of the BATRAC arm-training
machine. Movement time 2 (MT2) was defined as the movement onset from the target
back to the start position. 2. Time to peak velocity (TPV) was a measure of absolute
time measured in seconds from the point of movement onset to peak velocity.
3. Peak velocity (PV) corresponds to a moment in time when the highest velocity is
reached where acceleration is at or near zero at the changeover from the acceleration to
the deceleration phase. PV is calculated from the rate of change over time.
4. Acceleration was calculated from the slope or inclination of the velocity curve and
includes the percent of reach within this curve. Acceleration corresponds with the time
period of movement onset to peak velocity.
Posteriori of velocity profiles
The following metrics were analyzed posteriori to assess the movement smoothness
of the velocity profiles during the novel reaching task pre- and post-test.
1. Jerk metric was calculated from the average value of the absolute jerk divided
by the peak velocity of the corresponding trial (Jerk is defined as the rate of change of
acceleration). Jerk metric is assigned a negative value so as the smoothness increases, the
jerk metric also increases.
2. Speed metric was calculated from the average velocity divided by the peak
velocity of the corresponding trial. As the smoothness increases, speed metric also
3. Movement arrest period ratio (MAPR) was the amount of time that the
velocity profile was less than 10% of the peak velocity divided by total time of the trial.
A smaller MAPR indicates a smoother velocity profile.
4. Peaks metric was the number of peaks in the velocity profile greater than
0.5m/s multiplied by negative 1. As the smoothness increases, the peaks metric increases.
5. Tent metric was calculated from the area under the velocity profile curve
divided by the area of a curve draped over the top of it. The closer tent metric is to 1, the
smoother the velocity profile.
Data storage conformed to HIPAA regulations. VICON-captured data was stored
in a database on a secure network. Participants were assigned a participant number and
this number was the only identifier stored on these databases. A list of the participants'
names and participant numbers was kept in a locked file in Dr Lorie Richard's office.
Only the investigator and Dr Richards had access to this list. Only study personnel had
access to the participant notebooks or the database.
For each dependent variable a repeated measures 2 (time) x 2 (task) ANOVA was
performed with an alpha level of .05. Greenhouse-Geisser's adjustment of degrees of
freedom was applied to correct for small departures from the assumption of normality
and equality of variance in the two-factor design. A Bonferroni correction factor was
used to correct for multiple analyzes for the primary and secondary dependent variables.
The corrected alpha level for the primary dependent variables was .05, secondary
dependent variables was .01, and .01 for the posteriori analysis of the metrics. Lastly, a
descriptive analysis of the individual data was performed post-hoc. The sample of this
study was made up of a heterogeneous group of individuals with various levels of
severity at baseline; therefore we decided to examine the data descriptively at the
individual level, conjecturing that this could provide information on subgroups of
individuals that may have benefited, but would not have been detected in the group
Kinematic outcomes were analyzed with a repeated measure ANOVA for each of
the primary dependent variables. The within subjects factors were time (pre-test/post-test)
and task condition (novel task #1 and novel task #2). The primary dependent variable was
hand path trajectory. The secondary dependent variables were: 1) movement time, 2) time
to peak velocity, 4) peak velocity and 5) the percentage of the reach that was
An alpha value was set at .05 and corrected with Bonferroni to account for multiple
analyses on the secondary and posteriori analysis. The corrected alpha levels on
dependent variables were .01 for the secondary and the posteriori analysis of the velocity
profiles. Individual data will be reported last to identify individual differences that may
account for a pattern in the data.
Primary Spatial Dependent Variable
Table 3-1 displays the means and standard deviations for all outcome measures for
primary and secondary dependent variables. Table 3-2 displays the ANOVA summary
Hypothesis la and lb HPT will become straighter after intervention for novel task
#1 (generalization of similar task). No change will be noted for HPT on novel task #2 (no
generalization to dissimilar task).
Hand path trajectory
HPT was calculated as the ratio of the actual path traveled to the ideal straight line
joining the initial and final positions for the index finger or metacarpalphalangeal joint of
the 2nd digit (when subjects could not extend the index finger). There was a significant
main effect of time for HPT with hand path trajectories straighter at posttest (Figure 3-1).
However, there was no main effect of task nor interaction between the two variables. The
increase in straightness was found for both novel task #1 and novel task #2. Therefore
control of the spatial parameters of reach gained with intervention generalized to both
similar and dissimilar tasks.
Secondary Temporal Dependent Variables
The main effects are reported below and means and sd are displayed in Table 3-1. The
ANOVA summary table for the secondary dependent variables is displayed in Table 3-3.
Hypothesis 2a and 2b Generalization to novel task #1 following intervention will
be evident with decreased movement times. No improvement in movement times will be
demonstrated on novel task #2.
Contrary to expectations, the time to touch the target (MT1) was not significantly
different post compared to pre-intervention, nor across tasks. MT2, although shorter for
novel task #1 compared to novel task #2, also did not change with intervention
(Figure 3-2). There were no significant interaction effects of time and task on either
The data for MT1 and MT2 did not support the hypothesis that there would be
improvement in motor control on an untrained task with similar joint angles to the
training task. Movement time at post-test was no shorter than at pre-test. Bilateral
training did not appear to improve motor control during the performance of a task that
was similar in terms of joint angles to the training task. Not surprisingly, no
improvements in movement time were found for novel task #2, a task with dissimilar
joint angles to the training task which was in support of the hypothesis.
Hypothesis 2a and 2b TPV will increase for novel task #1 following intervention:
TPV will not change with novel task #2.
Time to peak velocity
TPV was measured from the point of movement onset to peak velocity. There was
no difference in TPV across time or tasks. There were no significant interaction effects.
Therefore, the training had no effect on the TPV.
Hypothesis 2a and 2b: PV will increase for novel task #1 and will not change for
novel task #2 following 2 weeks of intervention.
PV was the highest velocity that occurred during the reach. It typically occurs at the
moment of changeover from acceleration to deceleration in reaching to a target. There
was a significant main effect of time for PV (Figure 3-3) with PV larger following
intervention for both novel tasks. There was no main effect of task nor interaction
between time and task. Therefore, generalization of training was seen for similar and
Hypothesis 2a and 2b The acceleration phase of the reach will approach 50% of
the curve for novel task #1 and will not change for novel task #2 following 2 weeks of
There were no main effects or interaction effects noted for percentage of the reach
Posteriori of velocity profiles
Although this analysis was performed posteriori, changes in the smoothness of the
velocity curve could be beneficial in interpreting data collected on the reaching pattern of
individuals post stroke. An ANOVA summary table for the post hoc analysis is displayed
in Table 3-4.
Changes in the metrics of the velocity profile would point toward a smoothing of
the velocity curve following intervention indicating fewer stops and starts in the reaching
pattern toward the target. Smoothness of the curve would infer that the coordination of
the motor pattern for reaching has improved. There was a significant main effect of time
for the Peaks Metric with improvement post-intervention however no main or interaction
effects for task (Figure 3-4). In addition, there were no main or interaction effects for the
remaining smoothness metrics.
Descriptive Individual Data
Although this analysis is posteriori, this individual data could be beneficial in
planning future trials. Table 3-5 displays the individual data pattern of change across
primary and secondary variables for each subject.
Individual differences were analyzed comparing the raw score difference from pre-
to post-test with the pre-test sd for each dependent variable. Individuals who
demonstrated a change at post-test greater than the sd for the dependent variables at pre-
test are reported below. No subject improved in all variables with intervention.
Interestingly in three of fourteen subjects no change for any dependent variable was
noted and 4/14 subjects demonstrated a change in only one dependent variable. Five of
fourteen subjects demonstrated differences greater than the sd of the pre-test on three or
more variables. Thus, the descriptive individual data shows no general pattern across
subjects or subsets of subjects. Training did not frequently foster change in similar tasks
(novel task #1), but sometimes did in dissimilar tasks (novel task #2). In actuality change
was infrequent across the board.
Object 3-2. Testing conditions for generalization indicated some individuals changed.
Object 3-3. Testing conditions for generalization indicated some individuals changed
Table 3-1. Displays the means and standard deviations (sd) for task condition # 1 and #2
for all dependent variables pre- and post-intervention.
Table 3-2. Summary ANOVA model for the primary dependent variable. Significant p-
value = 0..05
Dependent variable after two weeks of intervention
Source Mean Square F (1,13) p-value
HPT Time .426 8.132 .014
Task 5.197 1.700 .215
Time*Task 9.072 .901 .360
Table 3-3. Summary of ANOVA model the secondary dependent variables. Significant p-
value = 0.01.
Dependent variables after two weeks of intervention
Source Mean Square F (1,13) p-value
MT 1 Time 4.48 2.248 .158
Task .614 .726 .410
Time*Task .260 .280 .606
MT 2 Time 6.84 1.897 .192
Task 3.14 8.167 .013
Time*Task .714 1.076 .319
TPV Time 1.491 .126 .728
Task 4.022 3.439 .086
Time*Task 1.143 .659 .427
PV Time 4.274 8.205 .013
Task .130 .212 .653
Time*Task .767 .767 .397
Accel Time 5.207 .392 .542
Task 6.864 1.114 .311
Time*Task 1.931 1.590 .229
Table 3-4. Summary ANOVA model for posteriori analysis
Significant p-value = 0.01.
Source Mean Square F (1,13)
of velocity profiles.
Table 3-5. Summary of individual change at post-test greater than the pre-test sd for each
dependent variable (x > sd)
Subjects FM-UE MT1 MT2
Task #1 24/28 X
Task #2 X X X
Task#1 51/49 X X X X
Task #2 X X X
Task #1 48/44
Task#1 33/38 X
Task #1 51/53
Task #1 46/45 X X X
Task #2 X
Task#1 37/38 X X
Task #2 X
Task #1 59/62
Task #2 X
Task #1 64/59
Task #1 52/61 X
Task #2 X X
Task #1 43/45 X
Task #2 X X
Task#1 35/34 X X X
Task #2 X
Task #1 18/23
Task #2 X X
Task #2 X
Task 1 2/14(14%) 1/14 (7%) 1/14(7%) 5/14(36%) 3/14(21%) 2/14(14%) 2/14(14%)
Task 2 2/14(14%) 1/14(7%) 2/14(14%) 5/14(36%) 2/14(14%) 0/14 (0%) 4/14(29%)
Hand Path Trajectory
HPT pre-test 1 HPT post-test
Figure 3-1. Significant main effect displayed for HPT for post-test. A value of 1 equals a
Movement Time 2
1 MT2 novel task #2
Figure 3-2. Significant main effect displayed for MT2 novel task #1.
VTT2 novel task #1
PV pre-test 1 PV post-test
Figure 3-3. Significant main effect displayed for PV on the post-test.
I T -0.761
Figure 3-4. Significant main effect displayed for posteriori Peaks Metric post-test.
Generalization is the ability to perform similar motor skills that were not
specifically practiced as part of the training intervention. Generalization is directly related
to the amount of practice on a particular task and how much motor learning occurred
during the acquisition phase of the new the task. In addition, for neurologically intact
individuals learning specific motor patterns, generalization occurs only under highly
similar spatial conditions that have similar neural processing elements and demands.1-3
We do not know if this is also true for the basic coordination skills that persons relearn
after stroke. Persons post-stroke are just regaining the capacity to move and learning how
to perform tasks with a decreased CMN pool output often their movements are influenced
by pathological synergies. The critical components of invariant task features necessary
for generalization of a motor skill in individuals with stroke under these conditions have
not been clearly delineated in the literature. 1
The specific aim of this study was to test spatial generalization of repetitive
rhythmic bilateral training to two novel reaching tasks in individuals with stroke. The
training task was a set of repetitive continuous movement reversals constrained by the
BATRAC equipment which allowed only reaching forward in front of the body within a
limited range (80% of the available reach). The transfer task (novel task #1) had similar
neural demands for joint angles, workspace, visual gaze angles, and muscle timing
compared to the training task but was not performed on the BATRAC. The second
transfer task (novel task #2) had dissimilar joint angles, workspace, and visual gaze
angles to the training task. It was predicted that improvements in kinematic parameters of
movement gained through two weeks of intervention would generalize to the similar
novel task #1 at post-test but not to the dissimilar novel task #2. Generalization occurred
at post-test for HPT but surprisingly was equivalent across the similar and dissimilar
tasks.. Generalization of training was further supported posteriori by the Peaks Metric
that was also significant at post-intervention. The changes in dependent variables taken
together indicate improved coordination with decreased stops and starts in the novel
We also predicted that movement speed might also improve with training. In fact,
peak velocity was also significant at post-test, with similar changes noted for both testing
tasks. Perhaps, novel task #1 and novel task #2 did not present task features that were
dissimilar enough to delineate a change between them kinematically post-intervention
The remaining temporal dependent variables of MT1, MT2, TPV, and acceleration
did not generalize to either task. The lack of change in these particular temporal
parameters may have been because speed of movement was not emphasized in the
training or testing tasks. On the other hand, the higher PV suggests that for at least some
of the reach, movements were faster. Perhaps subjects moved faster during parts of the
reach, but slowed their movements in the remaining sections of the reach, resulting an
overall unchanged movement time. Only a more detailed analysis of the reaching strategy
would allow firm conclusion on this issue. Thus, the results of this study offer only weak
evidence that repetitive rhythmic bilateral training may generalize to untrained tasks.
Unfortunately there was no measure of actual intervention task learning in this study to
delineate the lack of motor learning from the lack of generalization. Instrumentation of
the BATRAC equipment would allow for measurement of the temporal coordination of
the UE's providing more accurate measures of motor learning on the training task. This
information could then be utilized to make further assumptions about motor learning of
the intervention task compared to generalization of the novel transfer task.
Neuroscience Rationale for Repetitive Rhythmic Bilateral Training
The rationale for the potential of the BATRAC as an UE training tool post-stroke
has been theorized to tap into basic neurophysiological mechanisms that stimulate
coordinative structures priming the nervous system and firing up the CMN pool. Studies
have shown activation of both hemispheres with bilateral movements that are organizing
in a tight phasic relationship.4-8The symmetrical temporal relationship between the
hemispheres during bilateral activities may assist in laying down a template for basic
movement components necessary for reaching. However, neither the specific movement
characteristics of bilateral practice necessary for generalization nor the level of severity
of individuals post-stroke that would be most responsive to this type of therapy have been
The inconclusive findings of this present study are in contrast with the preliminary
data regarding bilateral utilization presented in the literature.10' 11 Using the BATRAC,
Whitall et al.,10 demonstrated significant improvement in UE motor performance on the
FM and WMFT. The results in the Whitall et al.10 study could be interpreted as
generalization to untrained tasks since the items on these testing measures were not
trained specifically in the bilateral intervention.
Luft et al.11 showed that repetitive rhythmic bilateral training using the BATRAC
influences neural mechanisms underlying motor skill in a small number of subjects. They
found increased hemispheric activation during paretic arm movements after training.
Although the motor skill assessed by fMRI in this study was not a reaching task as was
the transfer task in the present study, increased activity of the contralesional hemisphere
was observed. The Luft et al.11 study suggests that the BATRAC intervention induces
reorganization of motor networks in persons post stroke and Whitall's10 work further
suggests that reaching post bilateral intervention may improve.
Differences in our subject population and the kinematic parameters selected as
dependent variables could account for the different results observed in this study. Whitall
et al.10 used the FM and the WMFT to measure efficacy of bilateral training. These
assessments are a composite of summary scores over multiple items. Looking more
closely at the items that make up the assessments reveals that some items increase and
some items do not, while other items may even decrease a little. Therefore the overall
score can show improvement, while individual items themselves may not. This study
focused on the kinematic measures of one single task only 2 tasks, one that was similar to
the bilateral training. It is hard to compare the results of performance on only 2 tasks
with summary scores on tests made up of many tasks. Perhaps had we chosen different
tasks, we would have also found improvement across our subject sample. The task that
was chosen demonstrated variability between subjects and overall did not show change
across the multiple dependent variables, perhaps a different task or set of tasks may have.
Ideally a combination of clinical, kinematic, and kinetic measures would give a more
complete understanding of movement behaviors. Therefore, it is difficult if not
impossible to compare the results of the Whitall et al.10-12 studies with the present study
due to the differences the nature of the outcome measures.
Factors Potentially Affecting the Study Results
Practice schedules are known to affect motor learning, and could possibly have
played a role in our results.13, 14 This study utilized a distributed model of practice as did
Whitall et al.10 Distributed practice allows for a rest period equal to or greater than the
intervention period. In this study the intervention period equaled 5 minutes and the rest
period 10 minutes. Studies by Lee and Genovese15 have demonstrated in healthy
individuals that transfer performance was increased for groups that had longer rest
periods versus work periods. Other studies have also shown that distributed practice has a
large positive effect on learning.16, 17
The intervention in this study was also delivered in a blocked manner (grouping
like trials together) however, while blocked practice improves acquisition of a task,
random practice appears to be superior for true learning: retention and generalization of
the skill when tested after the training period of an intervention.18-20 Although, Whitall et
al.10 demonstrated improved upper extremity functional measures with this type of
practice (distributed blocked practice) during bilateral training on the BATRAC. The
blocked practice schedule perhaps limited the degree of motor learning that occurred
during the intervention and therefore, limited generalization to motor skills not practiced
directly in the training. Introducing a distributed random practice schedule for the in/anti
phase trials or randomly changing the metronome frequency on trials may have enhanced
the amount of motor learning and thus generalization following this intervention.
Random practice schedules increase the degree of problem solving during execution of
the task by introducing variability in the practice and in turn enhance retention and
generalization ultimately improving the amount of motor learning.1
Providing practice in a more condensed format (2 weeks versus 6 weeks) may have
contributed to the study results. Whitall et al10-12 provided the same training but
distributed the practice over a longer period of time. Currently it is not known if
condensed practice offers differential benefits compared with more distributed practice.
Dettmers21 found that CIMT distributed over 3 weeks with a shorter trial per day ( 3
hours/day) demonstrated improved UE function and quality of life in persons post-stroke.
Page et a122 also showed improved motor skills with a modified form of CIMT provided
in a distributed fashion. However, no study has directly tested a condensed version
against a more distributed version of an identical therapy to determine whether such
practice distribution influences the amount of motor gains experienced in therapy. The
difference in the distribution of practice between this study and Whitall's0o may prove to
be a critical factor in the resultant study outcomes and should be directly investigated in
future research. The distribution and dose of practice to effect a change in motor learning
in individuals post-stroke is clearly not understood.
Several additional aspects of this training may not have been optimal for motor
learning. First, repetitive rhythmic bilateral training may not serve as a robust learning
model since problem solving and the development of a reference of correctness of
movement are not inherently strong or emphasized in the training. The environmental
constraints of the BATRAC (the track and chest restraint) and repetitive nature of the
practice may have resulted in low neural demands and little problem solving. The training
was performed on a track that guided the spatial trajectory of the movements
furthermore; sensory cueing was provided from a metronome, which was self-paced to a
comfortable speed for the participant. This auditory cueing set up a temporal template for
the individuals to match guiding the temporal parameters of the movements. Each end
point stop provided kinesthetic cueing assisting with the timing of the reversal and
indicated the proper extent of movement. Thus, the environment provided specific spatial
and temporal features with predictable consequences, minimizing demands for problem
Figure 4-1. BATRAC invariant features, (A) in-phase and (B) anti-phase.
Introducing a margin for potential errors during the intervention might allow for
increased problem solving during task learning which would facilitate the development of
the capability to produce more effective movements and the ability to assess ones own
movement behaviors. Error correction builds movement strategies and a larger repertoire
of available movements to accomplish a task. Schmidt 1 has suggested that these aspects
of motor learning are necessary to retrieve information from long-term memory for
executing a motor skill. Shadmehr23 would argue these components are critical for the
formation of an internal model that would represent the physical dynamics of the limb
and the workspace environment (where the motor skill is performed in relationship to the
body/trunk/ upper extremities). Errors experienced in the training would influence the
performance of the motor skill in an untrained but similar task, contributing to the
performance of the generalization task after intervention. 1
One could argue that learning to coordinate the two upper extremities for in-phase
and anti-phase movement patterns as well as matching the metronome beat was practice
that afforded some degree of problem solving with this intervention. In fact, observing
participants during intervention revealed that indeed there was some difficulty in
coordinating the two limbs to obtain and maintain the temporal phasing of the movement
patterns as well as temporal matching with the metronome. However, temporal phasing
and matching were often accomplished with manual guidance of the therapist which has
been shown to improve performance during the task but not improve motor learning.1' 3
Thus, the degree of problem solving and development of a reference of correctness
during the intervention may not have been sufficient enough for motor learning and
generalization to a novel untrained similar task. If this is true, it is not surprising that little
generalization was found.
Another potential factor affecting the results was that perhaps subjects were not
very engaged in the learning task. The nature of repetitive rhythmic bilateral training on
the BATRAC was similar to an exercise with multiple repetitions versus meaningful
functional task practice. Practicing real life tasks that are motivating to the individual has
been shown to improve acquisition of a motor skill. Wu et al.24demonstrated better
kinematic performance of reaching movements to real objects when compared to
movements without relevant objects in persons post-stroke. Nelson et al.25 studied the
effects of an occupationally embedded exercise on bilaterally assisted supination in
persons post-stroke. Significant results were found for the group receiving real life
(meaningful) practice compared to rote exercise of the same task. Because the BATRAC
lacked meaningful practice, the intervention may have held little motivation for the
participants. Perhaps the results of this intervention would be strengthened if such
bilateral, repetitive exercise were combined with functional task practice that provides
motivation and meaning for the participants for real life skills improving their motor
Generalization is bolstered when the invariant features are similar between the
training intervention and the novel transfer task. Invariant features of a motor task consist
of the unique traits and rules that are particular to that task. The components chosen as
critical in this study: similar neural demands, joint angles, workspace, angle of visual
gaze, and muscle and joint synergies, may not have been similar enough for
generalization to consistently occur across the selected kinematic outcome measures.
Motor learning, generalization, and retention of a task are strengthened when the training
is varied.1, 3The movement of the upper extremities on the BATRAC did not allow for
diversified movement and therefore there was little room for error detection and the
development of a reference of correction. This intervention involved little to no
interaction with the environment since the objects in the environment did not change
from one attempt to the next. The subject held a handle or was strapped to the handle on
the BATRAC, which offered some degree of weight bearing on the apparatus, and
therefore training motions were closed chain movements. The individual was assisted to
the end point stops if they were unable to complete the task independently. The trunk of
the individual was blocked from forward flexion by a chest plate and proximal arm
motion was encouraged to complete the task. The demands of the task were externally
paced, predictive; the movement was fixed within a particular range and repetitive in
nature requiring minimal monitoring by the participant. Although most subjects
attempted to stay in time with the metronome they were assisted with verbal cues and
occasional manual assistance to coordinate their movements with the beat of the
Novel transfer task #1 was chosen because it had many of the characteristics
thought to be important for generalization to occur. The workspace, visual gaze angles to
the target, and joint angles required during the test reaches were identical to the training
intervention utilizing similar muscle and joint synergies to perform the task (novel task
#2 was dissimilar in these parameters). Schmidt1 et al have demonstrated that these
components are necessary for a motor skill to generalize to a novel untrained task.1 While
the lack of generalization may have been due to a lack of learning altogether, the similar
transfer task had other features that were different from the intervention task, which
possibly contributed to the limited generalization.
The transfer task may have required different processing compared to the training
task. Unlike the training intervention there was no track or chest restraint utilized in the
generalization task. The transfer task was a discrete unilateral reaching movement made
in free space with the affected upper extremity to the target and back to the start position.
The only environmental sensory cues were visual in nature: the target and the start
position. No auditory cueing was used other than a verbal cue to begin the movement.
The transfer task required transport of the upper extremity through space against gravity
to the target and back to the start position. One could postulate that both transfer tasks
required greater muscle force to move the limb against gravity then was needed in the
training task although these data were not collected in this study. Certainly more
variations in the movement pattern were available as there were an increased number of
df at upper extremity joints as with the trunk in order to control the transfer task.
Controlling such increased df in the transfer task was quite different than during training
with the BATRAC as an environmental constraint and the chest plate limiting trunk
forward lean. Allowing diversified movement within the execution of the novel task may
have increased the requirements for attentional demands, neural control, problem solving,
and error correction which were not inherent in the intervention. The transfer task
required focused attention on the target, control of the upper extremity through space, and
proprioception (to avoid over or under reaching) unlike the training on the BATRAC
where movements were guided by the track and kinesthic cues were provided by the
Examination of the invariant features reveals differences in the intervention-
training task and the transfer task that may help to account for the lack of generalization
across the dependent variables. Moving the arm fully through space required controlling
many df which was clearly not something that was trained. Although some of the
environmental constraints were similar: workspace, visual gaze angles; the change from a
closed chain movement in the training intervention to an open chain movement and
increased movement possibilities on the transfer task introduced an increased demand on
the muscular and neural systems as well as the need to problem solve the execution of the
task against gravity.
Identifying the specific features of the repetitive rhythmic bilateral intervention that
will foster generalization to an untrained motor skill in stroke has yet to be determined.
The training may not have provided a strong enough training stimulus or the components
chosen as critical in this study: similar neural demands, joint angles, workspace, angle of
visual gaze, muscle and joint synergies may not have been similar enough for
generalization to consistently occur across the selected kinematic outcome measures. A
lack of complexity in this intervention in regard to decreased demands to problem solve
and lack of development of a reference of correctness may have influenced the amount of
motor learning and further restricted generalization of the intervention to a similar
Limitations to the Study
This study was composed of a heterogeneous group of subjects therefore the
statistical power was influenced by the variance between the subjects (Tables 3-2 and
3-3). A small sample size (n=14) makes it difficult to draw conclusions about the
population in the study and results in a substantial reduction in power. Multiple analyses
on dependent variables were performed but adjusted with a Bonferroni correction
resulting in a more conservative alpha value further reducing the power. Previous
bilateral studies have used small sample sizes and demonstrated significant results on
motor outcome measures but kinematic outcomes have not been reported. Calculating
power at .80 for subsequent studies of this type of bilateral training with the same
dependent variables would require 1050 subjects for significant results. To counter the
violations to the assumption of normality a larger sample size and homogenous sample
would be beneficial in subsequent studies. A homogenous sample in the stroke
population may be difficult to achieve due to the variability of the insult to the CNS.
The inclusion criteria for this study were broad including subjects that had only
palpable extrinsic forearm finger muscle activity and some active shoulder and elbow
motion. Although this intervention was focused on proximal joint and muscle effector
systems and did not train grasp, perhaps the criteria should be modified to include some
degree of finger and hand motion indicating more motor recruitment in the upper
extremity. The inclusion criteria also did not include the ability to move the UE against
gravity a specified degree, which was an inherent part of the transfer tasks and may have
affected the amount of generalization measured at posttest.
Some individuals may have improved with the intervention however delineating
the variables that would discriminate those who might improve from those who might not
improve were limited. Measures to assess multiple factors that potentially could
influence each subjects' baseline and post-testing should be considered: 1) strength in the
upper extremity, 2) coordinative patterns /muscle joint synergies during movement 3)
spasticity, 4) praxis 5) executive functioning, 6) motor learning style (spatial/temporal),
and 7) sensory/proprioceptive status of the affected limb. Collecting the above data of the
participants' capability would further help to assess which subjects were able to benefit
the most from this therapy. Although the side of the lesion was not used as exclusionary
criteria in this study, it has been documented that individuals with a left-sided lesion have
more difficulty with rhythm keeping. 26-29 However, McCombe Waller and
Whitall12demonstrated that persons post-stroke with left-sided lesions improved greater
than persons with right-sided lesions during intervention with the BATRAC. Use of a
metronome in the intervention training sessions may affect the results for particular
subjects by cueing a temporal pace or disadvantageous by creating interference during the
acquisition period of learning where abstract neural signals are transformed into long-
term memory for retrieval at a later time for execution.1' 23 In our study, nine subjects had
left-sided lesions and five had right-sided lesions (Table 2-1). Although no clear pattern
of interference or advantage was evident from the individual data this should be a
consideration in future studies. Assessing the issue of interference of the metronome or
other rhythm keeper used in interventions during the acquisition stage of learning may be
advantageous and revealing when evaluating the results.
Models of the relationship of the dependent variables to each other are not clearly
developed in the field of movement science for the typical population and are evolving
for the stroke population. The numbers of possible kinematic variables and lack of a
model make it difficult to pick just one or two that are expected to change or which
would be the most important for function. The reliability and validity of kinematic
measures in the stroke population has not been established and therefore the stability of
the measures has not been substantiated.30 There was no obvious pattern among
individuals in regard to severity of involvement that predicted which subjects would
improve and on what kinematic outcomes in this study. Several investigators have
theorized that smoothness of the velocity profile demonstrates improved coordination of
the reaching movement with a decrease in stops and starts in the pattern of reaching in
healthy individuals and persons post-stroke.31-34 This supposition would connect the
kinematic variables of PV, HPT, and smoothness metrics, all contributing to the
understanding of coordination and quality of reaching ability in persons post-stroke.30' 35
37 Investigation into the correlation among these kinematic variables related to
performance on reaching tasks might allow selection of the most sensitive measures and
the prediction in their change after intervention.
This study specifically tested generalization of repetitive rhythmic bilateral training
to a similar novel task rather than the overall efficacy or motor learning in upper
extremity function. The kinematic results suggest that at a basic level repetitive rhythmic
bilateral training in and of itself are not enough to effect a change in motor control,
specifically generalization to an untrained novel motor skill across multiple dependent
variables. The novel task may not have held enough of the invariant features of the
training task to truly test generalization of this intervention. The importance of task
analysis of the invariant task features defined for the intervention versus the transfer task
cannot be underestimated. Identification of the critical components of the invariant
features necessary for generalization in the stroke population has yet to be determined.
The adequacy of the intervention in providing an opportunity for motor learning
involving problem solving and the development of a reference of correction without the
connection of real life practice should be scrutinized further. Lastly, the distribution of
practice on a continuous task may affect motor learning and the resultant outcomes and
should be investigated to determine optimal dosage during interventions post-stroke.
LIST OF REFERENCES
1. American Heart Association. Know the facts/Get the Stats. Available at:
http://www.americanheart.org. (Last accessed, April 2006)
2. Mayo NE, Wood-Dauphinee S, Ahmed S, Gordon C, Higgins J, McEwen S,
Salbach N. Disablement following stroke. Disability and Rehabilitation. May-Jun
3. Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. Recovery of upper
extremity function in stroke patients: the Copenhagen Stroke Study. Archives of
Physical Medicine Rehabilitation. Apr 1994;75(4):394-398.
4. Duncan P, Studenski S, Richards L, Gollub S, Lai SM, Reker D, Perera S, Yates J,
Koch V, Rigler S, Johnson D. Randomized clinical trial of therapeutic exercise in
subacute stroke. Stroke. Sep 2003;34(9):2173-2180.
5. Chen R, Cohen LG, Hallett M. Nervous system reorganization following injury.
Neuroscience. 2002; 111(4):761-773.
6. Hallett M. Plasticity of the human motor cortex and recovery from stroke. Brain
Res Brain Res Rev. Oct 2001;36(2-3): 169-174.
7. Shumway-Cook A, Woollacott MH. Motor control: theory and practical
applications. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2001.
8. Kandel ER, Schwartz JH, Jessell TM. Principles of neural science. 4th ed. New
York McGraw-Hill, Health Professions Division, c2000.; 2000.
9. Schimidt R, Lee TD. Motor control and learning: a behavioral emphasis. 4th ed.
Champaign, IL: Human Kinetics; 2005.
10. Kelso JA. Phase transitions and critical behavior in human bimanual coordination.
Am JPhysiol. Jun 1984;246(6 Pt 2):R1000-1004.
11. Kelso JA. On the oscillatory basis of movement. Bulletin of the Psychonomic
12. Tanji J, Okano K, Sato KC. Neuronal activity in cortical motor areas related to
ipsilateral, contralateral, and bilateral digit movements of the monkey. Journal of
Neurophysiology. Jul 1988;60(1):325-343.
13. Debaere F, Wenderoth N, Sunaert S, Van Hecke P, Swinnen SP. Changes in brain
activation during the acquisition of a new bimanual coordination task.
14. Debaere F, Wenderoth N, Sunaert S, Van Hecke P, Swinnen SP. Cerebellar and
premotor function in bimanual coordination: parametric neural responses to
spatiotemporal complexity and cycling frequency. Neuroimage. Apr
15. Gerloff C, Andres FG. Bimanual coordination and interhemispheric interaction.
Acta Psychology (Amst). Jun 2002; 110(2-3):161-186.
16. Stinear JW, Byblow WD. Rhythmic bilateral movement training modulates
corticomotor excitability and enhances upper limb motricity poststroke: a pilot
study. JClin Neurophysiol. Mar-Apr 2004;21(2): 124-131.
17. Byblow WD, Lewis G, Stinear J, Carson RG. The modulation of excitability in
corticospinal pathways during rhythmic movement. In: Swinnen SP, Duysens, eds.
Neuro-behavioral determinants of interlimb coordination A multidisciplinary
approach. Boston: Kluwer Acdemic publishers; 2004:155-185.
18. Luft AR, McCombe-Waller S, Whitall J, Forrester LW, Macko R, Sorkin JD,
Schulz JB, Goldberg AP, Hanley DF. Repetitive bilateral arm training and motor
cortex activation in chronic stroke: a randomized controlled trial. Jama. Oct 20
19. Wyller TB, Sveen U, Sodring KM, Pettersen AM, Bautz-Holter E. Subjective well-
being one year after stroke. Clinical Rehabilitation. May 1997; 11(2):139-145.
20. Umphred D. Merging neurophysiologic approaches with contemporary theories:
setting the stage for discussion. In: Lister MJ, ed. Contemporary management of
motor control problems: proceedings of the II STEP conference. Alexandria, Va:
Foundation for Physical Therapy; 1991:127-140.
21. Bobath B. Treatment of adult hemiplegia. Physiotherapy. Oct 1977;63(10):310-
22. Brunnstrom S. Movement therapy in hemiplegia; a neurophysiological approach.
Philadelphia, PA: Harper and Row; 1970.
23. Adler S, Beckers D, Buck M. PNF in practice: an illustrated guide. 2nd ed. New
York: Springer; 2000.
24. Davies P, M. Steps to follow: the comprehensive treatment ofpatients in ih/
hemiplegia. 2nd ed: Springer; 2000.
25. Singer R. Motor learning and human performance. 3rd ed. New York: Macmillan;
26. Winstein CJ, Merians AS, Sullivan KJ. Motor learning after unilateral brain
damage. Neuropsychologia. Jul 1999;37(8):975-987.
27. Brown G, Burns S. The efficacy of neurodevelopmental treatment in paediatrics: A
systematic review. 2001.
28. Fetters L, Kluzik J. The effects of neurodevelopmental treatment versus practice on
the reaching of children with spastic cerebral palsy. Phys Ther. Apr
29. Salter J, Y C, LL P, LC M. Rehabilitation nursing approaches to cerebrovascular
accident: A comparison of two approaches. Rehabilitation Nursing. Mar-Apr
30. Waganaar R, Meijer 0, van Wieringen P, Kuik D, Hazenberg G, Lindeboom J,
Wichers F, Rijswijk H. The functional recovery of stroke: A comparison between
neuro-developmental treatment and the brunnstrom method. CScandJRehab Med.
31. Horak F. Assumptions underlying motor control for neurologic rehabilitation. In:
Lister MJ, ed. Contemporary management of motor control problems: proceeding
of the II STEP conference. Alexandria, Va: Foundation for Physical Therapy;
32. Cech D, Martin S, eds. Functional movement development across the life span. 2nd
ed. Philadelphia W.B. Saunders; 2002.
33. Nudo RJ, Wise BM, SiFuentes F, Milliken GW. Neural substrates for the effects of
rehabilitative training on motor recovery after ischemic infarct. Science. Jun 21
34. Cramer SC, Nelles G, Benson RR, Kaplan JD, Parker RA, Kwong KK, Kennedy
DN, Finklestein SP, Rosen BR. A functional MRI study of subjects recovered from
hemiparetic stroke. Stroke. Dec 1997;28(12):2518-2527.
35. Cao Y, D'Olhaberriague L, Vikingstad EM, Levine SR, Welch KM. Pilot study of
functional MRI to assess cerebral activation of motor function after poststroke
hemiparesis. Stroke. Jan 1998;29(1):112-122.
36. Matthews PM, Johansen-Berg H, Reddy H. Non-invasive mapping of brain
functions and brain recovery: Applying lessons from cognitive neuroscience to
neurorehabilitation. Restor Neurol Neurosci. 2004;22(3-5):245-260.
37. Donchin O, Cardoso de Oliveira S. Electrophysiological approaches to bimanual
coordination in primates. In: Swinnen S, Duysens J, eds. Neuro-behavioral
determinants of interlimb coordination: a multidisciplinary approach. Dordrecht,
The Netherlands: Kluwer Academic Publishers; 2004:131-153.
38. Liepert J, Bauder H, Wolfgang HR, Miltner WH, Taub E, Weiller C. Treatment-
induced cortical reorganization after stroke in humans. Stroke. Jun
39. Nudo RJ. Adaptive plasticity in motor cortex: implications for rehabilitation after
brain injury. JRehabilMed. May 2003(41 Suppl):7-10.
40. Nudo RJ. Functional and structural plasticity in motor cortex: implications for
stroke recovery. Phys MedRehabil Clin NAm. Feb 2003;14(1 Suppl):S57-76.
41. Friel KM, Heddings AA, Nudo RJ. Effects of postlesion experience on behavioral
recovery and neurophysiologic reorganization after cortical injury in primates.
Neurorehabilitation and Neural Repair. 2000; 14(3): 187-198.
42. Kleim JA, Barbay S, Nudo RJ. Functional reorganization of the rat motor cortex
following motor skill learning. Journal ofNeurophysiology. Dec 1998;80(6):3321-
43. Nudo RJ, Friel KM. Cortical plasticity after stroke: implications for rehabilitation.
Rev Neurol (Paris). 1999;155(9):713-717.
44. Calautti C, Baron JC. Functional neuroimaging studies of motor recovery after
stroke in adults: a review. Stroke. Jun 2003;34(6):1553-1566.
45. Kopp B, Kunkel A, Muhlnickel W, Villringer K, Taub E, Flor H. Plasticity in the
motor system related to therapy-induced improvement of movement after stroke.
Neuroreport. Mar 17 1999;10(4):807-810.
46. Schaechter JD, Kraft E, Hilliard TS, Dijkhuizen RM, Benner T, Finklestein SP,
Rosen BR, Cramer SC. Motor recovery and cortical reorganization after constraint-
induced movement therapy in stroke patients: a preliminary study. Neurorehabil
Neural Repair. Dec 2002; 16(4):326-338.
47. Taub E, Uswatte G, Morris DM. Improved motor recovery after stroke and massive
cortical reorganization following constraint-induced movement therapy. Phys Med
Rehabil Clin NAm. Feb 2003;14(1 Suppl):S77-91, ix.
48. Liepert J, Miltner WH, Bauder H, Sommer M, Dettmers C, Taub E, Weiller C.
Motor cortex plasticity during constraint-induced movement therapy in stroke
patients. Neuroscience Letters. Jun 26 1998;250(1):5-8.
49. Levy CE, Nichols DS, Schmalbrock PM, Keller P, Chakeres DW. Functional MRI
evidence of cortical reorganization in upper-limb stroke hemiplegia treated with
constraint-induced movement therapy. Am JPhys MedRehabil. Jan 2001;80(1):4-
50. Jang SH, Kim YH, Cho SH, Lee JH, Park JW, Kwon YH. Cortical reorganization
induced by task-oriented training in chronic hemiplegic stroke patients.
Neuroreport. Jan 20 2003;14(1):137-141.
51. Magill R, Hall K. A review of the contextual interference effect in motor skill
acquisition. Hum Mov Sci. 1990;9:241-289.
52. Winstein C. Designing practice for motor learning: Clinical implications.
Contemporary management of motor control problems: proceedings of the II STEP
conference. Alexandria, VA: Foundations in Physical Therapy; 1991:65-76.
53. Hanlon RE. Motor learning following unilateral stroke. Arch Phys MedRehabil.
54. Winstein CJ. Knowledge of results and motor learning--implications for physical
therapy. Phys Ther. Feb 1991;71(2):140-149.
55. Schmidt RA. Motor learning principles for physical therapy. Contemporary
management of motor control problems. proceedings of the II Step conference.
Alexandria, VA: Foundation for Physical Therapy; 1991:49-63.
56. Turton A, Fraser C, Flament D, Werner W, KMB B, Lemon RN. Organisation of
cortico-motoneuronal projections from the primary motor cortex: evidence for task-
related funciton in monkey and man. In: Thilmann, al e, eds. Spasticity:
Mechanisms andManagement. Berlin: Springer-Verlag; 1993:8-24.
57. Cauraugh JH, Kim SB. Stroke motor recovery: active neuromuscular stimulation
and repetitive practice schedules. Journal ofNeurology, Neurosurgery, and
Psychiatry. Nov 2003;74(11):1562-1566.
58. Cauraugh JH, Kim SB. Stroke motor recovery: active neuromuscular stimulation
and repetitive practice schedules. JNeurolNeurosurg Psychiatry. Nov
59. Sainburg RL, Ghez C, Kalakanis D. Intersegmental dynamics are controlled by
sequential anticipatory, error correction, and postural mechanisms. JNeurophysiol.
60. Shadmehr R, Moussavi ZM. Spatial generalization from learning dynamics of
reaching movements. JNeurosci. Oct 15 2000;20:7807-7815.
61. Hesse S, Schulte-Tigges G, Konrad M, Bardeleben A, Werner C. Robot-assisted
arm trainer for the passive and active practice of bilateral forearm and wrist
movements in hemiparetic subjects. Arch Phys Med Rehabil. Jun 2003;84(6):915-
62. Mudie MH, Matyas TA. Can simultaneous bilateral movement involve the
undamaged hemisphere in reconstruction of neural networks damaged by stroke?
Disability and Rehabilitation. Jan 10-20 2000;22(1-2):23-37.
63. Whitall J, McCombe Waller S, Silver KH, Macko RF. Repetitive bilateral arm
training with rhythmic auditory cueing improves motor function in chronic
hemiparetic stroke. Stroke. Oct 2000;31(10):2390-2395.
64. Platz T, Bock S, Prass K. Reduced skilfulness of arm motor behaviour among
motor stroke patients with good clinical recovery: does it indicate reduced
automaticity? Can it be improved by unilateral or bilateral training? A kinematic
motion analysis study. Neuropsychologia. 2001;39(7):687-698.
65. Stinear JW, Byblow WD. Disinhibition in the human motor cortex is enhanced by
synchronous upper limb movements. Journal ofPhysiology. Aug 15 2002;543(Pt
66. Kelso JA, Holt KG, Ruben P, Kugler P. Patterns of human interlimb coordination
emerge form the properties of non-linear limit cycle oscillatory process: theory
and data. Journal of Motor Behavior. 1981; 13:226-261.
67. Criscimagna-Hemminger SE, Donchin O, Gazzaniga MS, Shadmehr R. Learned
dynamics of reaching movements generalize from dominant to nondominant arm. J
Neurophysiol. Jan 2003;89(1):168-176.
68. Werhahn KJ, Conforto AB, Kadom N, Hallett M, Cohen LG. Contribution of the
ipsilateral motor cortex to recovery after chronic stroke. Ann Neurol. Oct
69. Temprado JJ, Swinnen SP. Dynamics of learning and transfer of muscular and
spatial relative phase in bimanual coordination: evidence for abstract directional
codes. Exp Brain Res. Jan 2005;160(2):180-188.
70. Welsh TN, Almeida QJ, Lee TD. The effect of postural stability and spatial
orientation of the upper limbs on interlimb coordination. Exp Brain Res. Mar
71. Sainburg RL, Wang J. Interlimb transfer of visuomotor rotations: independence of
direction and final position information. Exp Brain Res. Aug 2002; 145(4):437-447.
72. Sunderland A, Tinson DJ, Bradley EL, Fletcher D, Langton Hewer R, Wade DT.
Enhanced physical therapy improves recovery of arm function after stroke. A
randomised controlled trial. Journal ofNeurology, Neurosurgery, and Psychiatry.
73. Sterr A, Elbert T, Berthold I, Kolbel S, Rockstroh B, Taub E. Longer versus shorter
daily constraint-induced movement therapy of chronic hemiparesis: an exploratory
study. Archives ofPhysical Medicine and Rehabilitation. Oct 2002;83(10):1374-
74. Ouellette MM, LeBrasseur NK, Bean JF, Phillips E, Stein J, Frontera WR, Fielding
RA. High-intensity resistance training improves muscle strength, self-reported
function, and disability in long-term stroke survivors. Stroke. Apr 22
75. McCrea PH, Eng JJ, Hodgson AJ. Biomechanics of reaching: clinical implications
for individuals with acquired brain injury. DisabilRehabil. Jul 10 2002;24(10):534-
76. Cunningham CL, Stoykov ME, Walter CB. Bilateral facilitation of motor control in
chronic hemiplegia. Acta Psychol (Amst). Jun 2002;110(2-3):321-337.
77. Cirstea MC, Levin MF. Compensatory strategies for reaching in stroke. Brain. May
2000;123( Pt 5):940-953.
78. Cirstea MC, Mitnitski AB, Feldman AG, Levin MF. Interjoint coordination
dynamics during reaching in stroke. Experimental Brain Research. Aug
79. Rohrer B, Fasoli S, Krebs HI, Hughes R, Volpe B, Frontera WR, Stein J, Hogan N.
Movement smoothness changes during stroke recovery. JNeurosci. Sep 15
80. McCombe Waller S, Whitall J. Hand dominance and side of stroke affect
rehabilitation in chronic stroke. Clin Rehabil. Aug 2005;19(5):544-551.
81. Shea CH, Kohl RM. Specificity and variability of practice. Res Q Exerc Sport. Jun
82. Shea CH, Kohl RM. Composition of practice: influence on the retention of motor
skills. Res Q Exerc Sport. Jun 1991;62(2):187-195.
83. Lee TD, Genovese ED. Distribution of practice in motor skill acquisition: different
effects for discrete and continuous tasks. Res Q Exerc Sport. Mar 1989;60(1):59-
84. Ammons R. Acquisition of motor skill: III. Effects of initially distributed practice
on rotary pursuit performance. Psychology. 1950;40:777-787.
85. Reynolds B, Bilodeau I. Acquisition and retention of three psychomotor tests as a
function of distribution of practice during acquisition. Journal of Experimental
86. Tsutsui S, Lee TD, Hodges N. Contextual interference in learning new patterns of
bimanual coordination. Journal of Motor Behavior. 1998;30:151-157.
87. Newell KM, McDonald P. Practice: A serach for task solutions. American Academy
of Physical Education, Enhancing human performance in sport: New concepts and
developments. Vol 25. Champaign, IL: Human Kinetics; 1992:51-59.
88. Shadmehr R. Generalization as a behavioral window to the neural mechanisms of
learning internal models. Hum Mov Sci. Nov 2004;23(5):543-568.
89. Wu C, Trombly CA, Lin K, Tickle-Degnen L. A kinematic study of contextual
effects on reaching performance in persons with and without stroke: influences of
object availability. Arch Phys MedRehabil. Jan 2000;81(1):95-101.
90. Nelson DL, Konosky K, Fleharty K, Webb R, Newer K, Hazboun VP, Fontane C,
Licht BC. The effects of an occupationally embedded exercise on bilaterally
assisted supination in persons with hemiplegia. Am J Occup Ther. Sep
91. Alcock KJ, Wade D, Anslow P, Passingham RE. Pitch and timing abilities in adult
left-hemisphere-dysphasic and right-hemisphere-damaged subjects. Brain Lang.
Oct 15 2000;75(1):47-65.
92. Mavlov L. Amusia due to rhythm agnosia in a musician with left hemisphere
damage: a non-auditory supramodal defect. Cortex. Aug 1980;16(2):331-338.
93. Murayama J, Kashiwagi T, Kashiwagi A, Mimura M. Impaired pitch production
and preserved rhythm production in a right brain-damaged patient with amusia.
Brain Cogn. Oct 2004;56(1):36-42.
94. Vignolo LA. Music agnosia and auditory agnosia. Dissociations in stroke patients.
Ann N YAcad Sci. Nov 2003;999:50-57.
95. Krebs HI, Hogan N, Aisen ML, Volpe BT. Robot-aided neurorehabilitation. IEEE
Trans Rehabil Eng. Mar 1998;6(1):75-87.
96. Platz T, Denzler P, Kaden B, Mauritz KH. Motor learning after recovery from
hemiparesis. Neuropsychologia. Oct 1994;32(10):1209-1223.
97. Trombly CA, Wu CY. Effect of rehabilitation tasks on organization of movement
after stroke. Am JOccup Ther. Jul-Aug 1999;53(4):333-344.
Claudia Ann Rutter Senesac, PT, MHS, PCS, has over 27 years of pediatric clinical
experience. She is the owner and administrator of a pediatric physical therapy private
practice since 1984 and is a board certified clinical specialist in pediatrics. She received
her bachelor's degree in physical therapy and master's degree in health science from the
University of Florida. She is graduating with a Doctor of Philosophy in rehabilitation
science with her research interest focused on motor learning and motor control in
neurological and neuromuscular impaired populations: adult individuals who have
suffered a stroke, pediatric individuals that have suffered a SCI, cerebral palsy and
neuromuscular diseases. Investigations have included constraint induced movement
therapy, upper extremity intervention protocols for recovery, and locomotor training in
the pediatric population. She has been an adjunct faculty member of the Physical Therapy
Department at the University of Florida since 1979 and faculty Lecturer since 2003. Her
primary teaching responsibilities in the entry-level doctorate program include Functional
Anatomy I and II and Pediatrics in Physical Therapy.