Praxis and the right hemisphere

MISSING IMAGE

Material Information

Title:
Praxis and the right hemisphere
Physical Description:
ix, 133 leaves : ill. ; 29 cm.
Language:
English
Creator:
Maher, Lynn M., 1956-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1995.
Bibliography:
Includes bibliographical references (leaves 126-132).
Statement of Responsibility:
by Lynn M. Maher.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002060420
notis - AKP8500
oclc - 33884313
System ID:
AA00003608:00001

Full Text







PRAXIS AND THE RIGHT HEMISPHERE


BY

LYNN M. MAHER

















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


1995














ACKNOWLEDGMENTS


One of the benefits of having taken a long time to

complete this program is that I have been able to work with

many different individuals who have contributed to my growth

and development.

I am deeply indebted to my chair and mentor, Dr. Leslie

J. Gonzalez Rothi. Her teaching and guidance have had a

profound impact on my work, my career, and my life. Without

her patient persistence I surely would not have completed

this program. For her determination, conviction and

fortitude, I am eternally grateful.

I wish to express my sincere appreciation to Dr.

Kenneth M. Heilman for the gifts of his time, talent and

wisdom. I am extremely fortunate to have studied with

someone who is so willing to share his knowledge and

expertise. He taught me how to think and for that I am

thankful.

I wish to acknowledge the other members of my

committee, Dr. Russell Bauer, Dr. Lewis Goldstein and Dr.

Ira Fischler. It has been my privilege to be associated

with these exceptional teachers and I appreciate their

helpful guidance toward the completion of this dissertation.

ii






I acknowledge Lisa Lu and David Bandy for their

meticulous and herculean efforts as raters in this study. I

would also like to recognize those who participated in the

data collection used in this study: Dr. Cynthia

Ochipa, Dr. Margaret Greenwald, Dr. Mieke Verfaellie and Dr.

Linda Mack.

During my tenure at the University of Florida I have

had the good fortune to be surrounded by individuals who,

despite having no formal obligation to do so, have provided

much needed advice and encouragement. To Dr. Leilani Doty,

Dr. Dawn Bowers, Dr. Alice Holmes, and Dr. Tiana Leonard, I

extend my sincere thanks.

I would like to thank the Department of Neurology in

the College of Medicine for generously supporting the bulk

of my education and training. I wish extend my appreciation

to the members of the Neuropsychology Lab and the

Gainesville V.A. Audiology and Speech Department over the

past seven years. Special thanks go to Dr. D.J. Williamson,

Joy McCallum, Barbara Hawes, and Dr. Anastasia Raymer for

their assistance with this project. In addition, I wish to

thank the Atlanta V.A. Audiology and Speech Department for

granting me the time to complete this project.

A special thanks go to my good friends who have held

steadfast through the years; Cindy, Stacie, Mary, Angie,

Kate, Pat and Ken. I came to depend on them for strength as

well as friendship and was never disappointed.

iii






I would also like to thank Dr. Katherine Burge-Calloway

for her gentle advice and counsel.

I wish to thank my parents and my family for their

caring and encouragement. There were many times when I know

they were at a loss as to what they could say or do to help,

and I want them to know how much I appreciate their efforts.

I will always be thankful to Michael F. Maher for his

constancy, patience, compassion and love not only during

these years of study but throughout our lives together.

Finally, my love and appreciation go to Matthew Michael

Maher and Kathryn Lynn Maher, the only two people involved

in this endeavor who had no choice in the matter. I

dedicate this work to them in the hope that they will know

how much they mean to me.
















This research was supported by NIH Grant R01 NS 20204 and

NIH Grant R01 NS 25675 to Dr. Kenneth M. Heilman and Dr.

Leslie J. Gonzalez Rothi and by VA Rehabilitation Research

and Development Program Grant to Dr. Gonzalez Rothi.
iv






TABLE OF CONTENTS

ACKNOWLEDGEMENTS...... ........................ii

LIST OF TABLES.................................. vi

ABSTRACT...................................... viii

CHAPTERS

1 INTRODUCTION............................. 1

Definition............................ 5
Historical Perspective..... ...........5
Studies Supporting
Left Hemisphere Praxis Dominance...10
Praxis As A Distributed
Multidimensional System............15
Right Hemisphere Role in Praxis.......24
Assessment of Apraxia.................37
Questions of This Study...............40

2 METHODS.................................44

Procedures............................45
Analysis..............................59

3 RESULTS..................................63

Analysis of Group Characteristics..... 64
Research Questions....................67
Extraneous Variables..................78
Summary................................81

4 DISCUSSION ..............................82

Research Questions.....................83
Methodologic Issues/
Study Limitations................ 90
Clinical Implications..................96
Implications for Future Research......97
Summary................................. 98

APPENDICES

A LIST OF TOOLS AND AUDITORY COMMANDS..... 100
B CRITICAL GESTURE ELEMENTS...............102
C RATER TRAINING PACKET...................114
D BRAIN-DAMAGED SUBJECT LESION DATA.......124

REFERENCES....................................... 126

BIOGRAPHICAL SKETCH.............................133






LIST OF TABLES


Table Page

2-1 Subject Descriptive Information: Age.........48

2-2 Subject Descriptive Information: Education...48

2-3 Subject Descriptive Information: TPO..........49

2-4 Content Errors by Group......................55

2-5 Inter-rater Scoring Reliability ............. 57

2-6 Intra-rater Reliability......................58

2-7 Classification of Errors.....................61

3-1 Total Apraxia Score by Group................. 65

3-2 Internal Configuration Subtest Score.........65

3-3 External Configuration Subtest Score..........66

3-4 Occurrence Subtest Score.....................66

3-5 Movement Subtest Score.......................66

3-6 Amplitude Subtest Score......................66

3-7 Sequencing Subtest Score.....................66

3-8 Pairwise Comparison: Total Apraxia Score.....68

3-9 Kruskal-Wallis ANOVA for Gesture
Element Subtests............................70

3-10 Pairwise Comparison for Amplitude
and Movement ................................71

3-11 Total of Error Types on Internal
Configuration..............................74

3-12 Total of Error Types on Occurrence..........74

3-13 Total of Error Types on External
Configuration..............................74

3-14 Total of Error Types on Amplitude............74

3-15 Total Of Error Types on Movement............75

vi






3-16 Total of Error Types on Sequencing..........75

3-17 Index of Embellishment by Group.............79

3-18 Number of Location Cues by Group............80


vii






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



PRAXIS AND THE RIGHT HEMISPHERE


BY


Lynn M. Maher


August, 1995


Chair: Leslie J. Gonzalez Rothi
Major Department: Communication Processes and Disorders


Whereas in right handers it has been demonstrated that

the left hemisphere plays a critical role in the programming

of skilled movement, the role of the right hemisphere

remains unclear. The purpose of this study was to determine

if the right hemisphere (nondominant hemisphere) contributes

to the production of skilled limb movement in right handers.

The production of gestures by right hemisphere brain-damaged

subjects was compared with that of left hemisphere brain-

damaged subjects and non-brain-damaged control subjects

gesturing with their right or left hand on a task of gesture

to auditory command. The gestures were scored across six

dimensions yielding six individual gesture element scores,

each single element believed to be critical for the

production of skilled movement. The six scores were summed

to yield a total apraxia score. Comparing the right

hemisphere brain-damaged group with the normal controls

viii






using their right hand on the total apraxia score and on the

individual gesture element scores, it would appear that the

right hemisphere brain-damaged group overall did not display

a significant apraxia. The one error of gesture performance

that was noted in the right hemisphere brain-damaged group

when compared to the right hand normal controls was

amplitude. In contrast, when compared to the other

experimental groups, the left hemisphere brain-damaged group

made more movement errors. Descriptive error pattern

analysis revealed that in the production of the skilled

movement the left hemisphere brain-damaged group tended to

make more errors with the selection, activation and

stabilization of joints.












CHAPTER 1

INTRODUCTION

Ideomotor apraxia is a disorder of skilled movement

which cannot be attributed to elemental sensory or motor

deficits. While it has been a consistent finding that

lesions of the left cerebral hemisphere result in

significantly worse performance on praxis testing than

lesions of the right cerebral hemisphere (Liepmann and

Maas,1907; Geschwind, 1965; Goodglass and Kaplan, 1963;

Kimura and Archibald, 1974; De Renzi, Pieczuro and Vignolo,

1968), there has been some indication in the literature of a

right hemisphere contribution to the performance of skilled

motor behavior (De Renzi, 1985; Haaland and Flaherty, 1984).

Haaland and Flaherty studied left hemisphere and right

hemisphere brain-damaged subjects and their performance on

non-meaningful, intransitive and transitive gestures. While

there were praxis performance differences between right

hemisphere and left hemisphere brain-damaged groups in their

study, both brain-damaged groups performed worse than the

normal controls on all three types of gestures. The types

of errors made by the left hemisphere and right hemisphere

brain-damaged groups were not significantly different for

the non-meaningful and intransitive gestures, and both








brain-damaged groups made more partial errors than their

control counterparts. However, on the transitive gesture

production task, group differences in error type emerged.

The fact that Haaland and Flaherty (1984) were able to

find differences on praxis performance between right

hemisphere brain-damaged patients and their right-handed

normal controls raises the question of whether the right

hemisphere contributes to the performance of skilled motor

behavior. If praxis is the sole province of the left

hemisphere, one would predict normal performance in right

hemisphere brain-damaged patients as has been suggested in

the past. The failure to find more elaborate and specific

error pattern differences between left hemisphere and right

hemisphere brain damaged patients may indicate that the

information provided by both hemispheres is relatively

comparable. Alternatively, the lack of an error type

difference by right-hemisphere versus left-hemisphere brain

damaged patients may reflect the inability of the particular

scoring system used in the Haaland and Flaherty study to

capture the nature of the differences.

Rothi, Mack, Verfaellie, Brown and Heilman, (1988)

stress that any system used to describe the errors in praxis

performance must reflect the temporal and spatial nature of

limb movement. Rothi and her colleagues modified a scoring

system that had been developed by Klima and Bellugi (1979)

to capture the structure of the American Sign Language of








the Deaf. After applying this modified error typing system

to the analysis of praxis production of left hemisphere

apraxic patients and normal controls, Rothi et al. (1988)

identified six error types that distinguished apraxic

patients from normal controls. However, Rothi et al. did

not include right hemisphere brain-damaged patients in their

study. Thus it is still unknown if right hemisphere brain-

damaged patients would score differently than left

hemisphere brain-damaged patients using the scoring system

of Rothi et al. (1988). Furthermore, if Haaland and

Flaherty (1984) are correct in suggesting that the right

hemisphere contributes at least in part to the performance

of skilled motor movement, the nature of that contribution

is still unknown. Is the information provided by the right

hemisphere unique, or is it similar to the information that

is provided by the left hemisphere? If the right hemisphere

contributes something different than the left hemisphere to

praxis processing, then one might predict a differential

praxis error pattern would emerge as the result of right

hemisphere versus left hemisphere brain damage.

The specific objective of the proposed study is to

determine what contribution, if any, the right hemisphere

makes to the performance of skilled movement by applying a

scoring system that has been demonstrated to be sensitive to

the temporal-spatial aspects of praxis to the performance of

right-handed, right hemisphere brain-damaged patients on








praxis tasks. This error type analysis will be compared to

an error analysis of the praxis performance of right-handed

left hemisphere brain-damaged patients and to that of right-

handed non-brain-damaged (normal) controls. If the right

hemisphere is not involved in the production of skilled limb

movements, then we would predict that there would be no

difference between the right hemisphere brain-damaged group

and the normal controls in the error frequencies and

distribution of error types in their praxis performances.

If the right hemisphere does make a contribution to praxis

production, we predict that the right hemisphere brain-

damaged group will differ from the normal controls in the

error frequencies in their praxis performances. In this

case, if the contribution the right hemisphere makes to the

processing of skilled limb movement is unique, we would

predict that the right hemisphere brain-damaged patients

would perform differently than both the left hemisphere

brain-damaged patients and the normal controls with respect

to the distribution of error types in their praxis

productions. If the left hemisphere and right hemisphere

contribute the same information to the processing of skilled

limb movements, we would expect comparable performances

between left hemisphere and right hemisphere brain-damaged

patients in the distribution of error types in their praxis

performances, but differences in error frequencies when both

brain-damaged groups were compared to normal controls.








Definition

Ideomotor apraxia is a disorder of skilled movement

which cannot be attributed to other neurological causes of

motor impairment such as weakness, akinesia,

deafferentation, abnormal tone or posture, movement

disorders, intellectual impairment, impaired comprehension,

inattention, or uncooperativeness (Heilman and Rothi, 1985).

While the term apraxia has been applied to a variety of

behavioral disorders including dressing apraxia, buccofacial

apraxia, gait apraxia, constructional apraxia and swallowing

apraxia to name a few, we will limit this study specifically

to limb apraxia. The original description by Liepmann

(1900) defined apraxia as "the inability to act, i.e. to

move the moveable parts of the body in a purposeful manner,

though motility is preserved" (p.160). Though Liepmann was

apparently not the first to identify the deficit of skilled

movement or an inability to use tools (see Brown, 1972), he

was the first to study the disorder systematically, to

isolate the disorder in its pure form, and to propose a

neurological mechanism to explain the disorder.

Historical Perspective

Theory of Left Hemisphere Dominance for Limb Movement

Liepmann (1900) reported the case of a 48-year-old

government official who subsequent to a stroke presented

with unilateral apraxia. Originally he was considered to be

severely demented and aphasic until it was discovered that






6

his failure to respond appropriately to commands was related

not to a comprehension deficit, but to an inability to

perform with his right hand. When his right hand was

restrained, the patient was able to use his left hand to

complete desired responses. The striking disparity between

right limb and left limb performance for both upper and

lower extremities was evident on pointing tasks, imitation

tasks, and actual tool use. Object agnosia and auditory

comprehension deficits were ruled out as explanations for

response failure as the patient readily and accurately

responded with his left hand. Liepmann's emphasis in

reporting this case was to distinguish ideomotor apraxia

from disorders of visual recognition and auditory

comprehension.

In 1905, Liepmann proposed that the left hemisphere, in

addition to being dominant for speech performance, was also

dominant for action. After studying 89 brain damaged

subjects with evidence of either right hemisphere or left

hemisphere lesions, Liepmann reported that the origin of

"purposive movements i.e. those learned connections of

elementary muscle actions which either represent effects on

the object world or are manifestations of mental

events to others" (p.19) was in the memories for the

movements stored in the left hemisphere. The evidence for

left hemisphere dominance for action was found in the

observation that out of 42 patients with right hemisphere








brain damage, only "rarely" was there any difficulty with

gesturing actions, with copying motions or with actual tool

use. Of the 41 left hemisphere brain-damaged patients

reported in this study, approximately half of them

demonstrated difficulty on these tasks using their left limb

in the case of right sided hemiplegia, or bilaterally, in

those few cases that were not hemiparetic. Thus the

inability of the left limb to follow commands subsequent to

left hemisphere damage suggests that the right hemisphere is

dependent on the left hemisphere to some extent in producing

these movements.

Liepmann (1905) further divided the apraxic group into

two groups based on the presence or absence of aphasia,

which he used as his indicator of supracapsular lesions

(cortical lesions above the internal capsule). He observed

that the incidence of apraxia was far greater in the

supracapsular group. The praxis deficit could not be

dismissed on the basis of poor auditory comprehension, as

the patients appeared to be attempting the target and also

indicated their ability to comprehend on other tasks.

Furthermore, these patients also failed on the imitation of

movements which should be unaffected by auditory

comprehension. Liepmann observed that while the practice

disturbance seemed the result of an inability to produce

these learned movements from memory, it also manifest itself

on imitation tasks and therefore could not be considered a








memory deficit per se. He proposed that apraxia was the

result of a defect in the "space-time concept" of movement

and/or an inability of "praxis innervation" (p.28). The

disturbance was not a result of damage to the left hand

center itself nor to damage to Broca's area. Rather, the

deficit was attributable to disruption of the white-matter

fibers arising from the left hemisphere which project to the

right hemisphere. For a portion of the subjects, this

disturbance extended to the inability to use actual tools.

This study led Liepmann (1905) to propose the

following theory of learned action: As a result of practice

we develop "movement formula" or the knowledge of the space-

time sequence for action. These formula typically involve

visual information, but may also involve other sensory

modalities as well. As information is processed, these

formula are transformed into innervatory patterns which

allow for the execution of movement. For certain short,

repetitive and stereotyped movements, a kinesthetic memory

is also developed which allows for bypassing the innervatory

patterns. The movement formula are insufficiently

represented in the right hemisphere, and the adequate

performance of these movements by the left limb is dependent

on the memories stored in the left hemisphere. Thus, a

supracapsular lesion in the left hemisphere, while

frequently causing right sided hemiplegia, will result in a

left sided apraxia. The preserved ability of the right








hemisphere to produce some movements (albeit lacking in

sufficient sharpness and more difficult in the absence of

the tactile, visual and kinesthetic cues of actual tool use)

was stated to be the result of the residual kinetic

abilities of the right senso-motorium.

Liepmann and Maas (1907) later reported on a patient

(Ochs) with an acute onset of right hemiplegia, reduced

spontaneous speech production, agraphia, and an inability to

complete certain tasks with his left hand including actual

tool use. The patient was found to have adequate auditory

comprehension and gesture recognition, and was proposed to

demonstrate a "pure agraphia" resulting from his praxis

deficit (that is, not linguistic in nature) which was

labeled apraxicc agraphia." On autopsy the patient was

revealed to have softening in the left hemisphere in the

distribution of the anterior cerebral artery. The lesion

involved the white matter from the upper frontal convolution

and the origin of the callosal fibers making up the anterior

3/4 to 4/5 section of the corpus callosum. This resulted in

an isolation of the left hemisphere from the right

hemisphere except for the regions connected by the posterior

section of the corpus callosum. The right hemisphere was

intact. This case provided further support for Liepmann's

(1907) proposal that the left hemisphere is dominant for the

programming of "goal-oriented movements" of either hand.

Liepmann (1905) identified a distinction between the

production of action and the ability to comprehend the








meaning of an action, as many of the cases he described

could not produce gestures but could comprehend their

meaning. He seemed to suggest that there was a particular

area of the left hemisphere that was critical for the

concept of the action as exemplified in his description of

three cases which differed from the ideomotor patients in

that their copying of gesture was either normal or near

normal. He proposed that their deficit stemmed from an

inability to "arouse" the time-space-form picture of the

movement, rather than with the actual innervation of the

movement, i.e. an ideational apraxia rather than an

ideomotor one. He used these cases to support his notion

that the more posterior the lesion, the more likely the

deficit was to be a well-formed interchanging of movements

as opposed to a poorly produced movement.



Studies Supporting Left Hemisphere Praxis Dominance

During the seventy-five years following Liepmann's

seminal works most of what was reported was a confirmation

of his basic tenants of left hemisphere dominance for

skilled movement, the concept of the gesture memories, and

callosal transfer of skilled movement information from the

left hemisphere to the right motor cortex. The evidence in

support of Liepmann's hypotheses was found in three major

areas: l)the incidence of apraxia following left hemisphere

brain-damaged (hereafter referred to as LBD) versus right








hemisphere brain-damaged (hereafter referred to as RBD)

and/or controls with no history of brain damage (hereafter

referred to as NC), 2)the differences in motor skill

performance and in motor learning performance between LBD

patients and RBD patients, and 3)the differences between

left hemisphere versus right hemisphere motor performance

in normals.

In general, most of the early group studies in praxis

were designed to determine the occurrence of ideomotor

apraxia in samples of unilateral brain damaged patients.

For example, De Renzi et al. (1968) assessed ideomotor

apraxia in a sample of 160 LBD and 45 RBD patients. In this

sample, 45 LBD patients were identified as having ideomotor

apraxia, whereas out of 45 RBD patients none were ideomotor

apraxic. In a later study by De Renzi and his group (De

Renzi, Motti and Nichelli, 1980), they found the incidence

of apraxia to be 80% of LBD patients and 20% of RBD

patients. The findings of De Renzi et al. of a higher

incidence of ideomotor apraxia following left hemisphere

brain damage is consistent with that of Goodglass and Kaplan

(1963) who compared the praxis performances of groups of

LBD aphasic, non-aphasic RBD and bilateral lesioned patients

and concluded that LBD patients were more impaired on their

praxis testing than RBD patients. The importance of the

left hemisphere for skilled movement was also supported by

Kolb and Milner (1981). They compared patients with left








hemisphere and right hemisphere frontal and parietal

excisions on their ability to copy meaningless hand gestures

and found that those patients with left hemisphere parietal

lesions were significantly more impaired than patients with

right hemisphere parietal lesions or left or right

hemisphere frontal lesions. Likewise, Lehmkuhl, Poeck and

Willmes (1983) found very few praxis errors in their RBD

group and normal controls, but found a substantial deficit

in their LBD group. Further support for the hypothesis of

left hemisphere dominance for praxis was found in a report

by Alexander, Baker, Naeser, Kaplan and Naeser (1992) who

also compared LBD patients, RBD and normal controls on

praxis tests. They reported that the RBD group did not

differ from the normal control group in their praxis

performance. However, both anterior and posterior left-

hemisphere lesions resulted in apraxic performance.

Kimura and Archibald (1974) studied differences between

LBD and RBD patients on a variety of skilled movements

including finger control, imitation of static and dynamic

hand positions, transitive gestures to verbal command (i.e.

pantomime of tool use on an imagined object), intransitive

gestures to verbal command (i.e. gestures that do not

require a tool or an object to be acted upon, such as "wave

goodbye") and movement recognition. They reported that LBD

patients were more impaired than RBD patients when imitating

sequenced movements, including intransitive and transitive






13

gestures. Kimura et al. concluded that the left hemisphere

was critical for producing complex motor sequences.

Haaland, Harrington and Yeo (1987) also found evidence for

greater difficulty in the production of "open loop"

movements (i.e. movements requiring visual feedback and

therefore more like complex skilled movement) among their

LBD group relative to their RBD group and/or normal

controls. Finally, Wyke (1967, 1971) reported that patients

with LBD demonstrated bilateral deficits in generating

rapidly sequenced movements, whereas the RBD groups

demonstrated slowing of the contralateral limb only.

Left hemisphere versus right hemisphere differences

subsequent to brain damage have been described with respect

to motor learning (Kimura, 1977; Heilman, Schwartz and

Geschwind, 1975; Rothi and Heilman, 1985). Kimura (1977)

described deficits in the acquisition of sequenced motor

skills in LBD patients when compared to RBD patients. The

LBD group consistently made more perseverative errors and

also produced more unrelated movements than the RBD group.

Heilman, Schwartz and Geschwind (1975) demonstrated that LBD

patients with ideomotor apraxia had greater difficulty

learning a rotary pursuit task than LBD patients without

apraxia. Furthermore, the importance of the left hemisphere

praxis system for learning motor skills was supported by

Rothi and Heilman (1985) who demonstrated that LBD apraxic

patients had greater difficulty learning gestures than non-








apraxic LBD patients or normal controls. The apraxic

patients took longer to acquire gestures and learned fewer

gestures overall than the control subjects.

Differences in left hemisphere versus right hemisphere

processing that suggest left hemisphere dominance for motor

programming have also been observed in studies of normals.

For example, Taylor and Heilman (1980) reported more rapid

acquisition of a novel, nonverbally coded motor skill

following right hand training rather than left hand

training, suggesting a superiority of the left hemisphere

for motor programming. Balfour, Clarke and Geffen (1991)

used reaction time to detect differences between left and

right hand performance which increased with task complexity

in normals. They found laterality effects between the two

hemispheres that reflected differences in motor planning

(i.e. response preparation), but not in motor execution.

The authors concluded that sequential ordering of response

sequences is performed by the left hemisphere.

To summarize, the literature cited thus far comes from

a variety of sources including studies of normals and

studies of brain-damaged individuals. In addition, a

variety of techniques such as behavioral analysis of

pantomime and meaningless gestures, reaction time and rotary

pursuit have been reported evaluating not only praxis

performance but praxis learning as well. Collectively this

converging evidence supports Liepmann's hypothesis






15

suggesting that the left cerebral hemisphere has a distinct

role in praxis production. The next section will review the

evidence that suggests which brain region (or regions)

within the left hemisphere are implicated in praxis.

Praxis As A Distributed Multidimensional System

Liepmann's proposal of callosal disconnection resulting

in a unilateral ideomotor apraxia was substantiated by

Geschwind (1965). Geschwind and Kaplan (1962) described a

case study of a patient with a large callosal lesion

yielding an apraxia of the left limb (including writing)

which improved on imitation and tool use, while his right

hand was normal. Geschwind and Kaplan suggested that the

patient's deficits arose from a disconnection of the right

motor cortex from the speech areas (Geschwind, p.605).

Geschwind (1965) presented a model of gesture production

that resembled Wernicke's model of language processing

(Heilman and Rothi, 1993). That is, he suggested that

auditory input is processed in the left hemisphere by

Heschl's gyrus and Wernicke's area and the pre-motor cortex

and primary motor cortex via the arcuate fasciculus. Damage

to the arcuate fasciculus would result in a verbo-motor

disconnection between auditory processing and motor output.

Lesions deep to the supramarginal gyrus would also

disconnect fibers from the visual association and

somasthetic areas, accounting for the apraxic performance on

imitation and with actual tool use in some patients. The








sparing of these fibers would explain those cases who were

able to imitate and use tools (e.g. Geschwind and Kaplan,

1962). Geschwind (1965) also proposed that area 6 in the

left hemisphere was the association cortex for the motor

system and as such, callosal connections to the right motor

system would originate in this area (p.614). Lesions to

these fibers would disconnect left hemisphere motor pathways

from the motor program and would prevent callosal transfer

of the information to the right hemisphere, resulting in

bilateral apraxia. Lesions in the callosum itself would

result in a unilateral apraxia of the left limb.

Further support for callosal disconnection in apraxia

was provided in a case report by Watson and Heilman (1983).

They described a 43-year-old woman with an acute onset of

apraxia and apraxic agraphia which was confined to the left

hand. Praxis and writing performance with the right hand was

flawless, and recognition of gesture was intact. Acutely

post onset, the patient was unable to approximate the

gesture target with her left hand. She appeared to not have

any idea about the nature or use of tools. It must be noted

that the concepts of what these tools were used for were not

actually destroyed, as evidenced by her flawless performance

with the right hand and her ability to recognize gesture.

The "conceptual" deficit described by the authors was

explained as a disconnection of the concepts of tool use

from the right hemisphere. Unlike the Geschwind and Kaplan








(1962) case, this patient's performance did not improve on

imitation and with actual tool use. This suggests the

relative dependence of the right hemisphere in this patient

on the visuokinesthetic engrams of the left hemisphere. The

patient's CT scan revealed an extensive lesion of the body

of the corpus callosum, from the genu to the posterior one-

fourth to one-fifth of the body, leaving the splenium

intact.

Heilman (1979) refined Liepmann's concept of "movement

formula," referring to them as "visuokinesthetic motor

engrams." He proposed that these visuokinesthetic motor

engrams are stored in the dominant parietal cortex and are

used to program the motor association cortex with the

necessary time-space components to produce the action. This

schema suggests that ideomotor apraxia may arise from more

than one cause and that the nature of the deficit will

differ depending upon the location of the lesion. The

patient with a disconnection syndrome (as described above)

should be able to judge and recognize gestures accurately,

because of the preservation of the visuokinesthetic motor

engrams in the left parietal lobe. The patient with damage

to the engrams themselves, however, should have difficulty

recognizing the distinguishing features of a particular

action. To test this hypothesis, Heilman, Rothi and

Valenstein (1982) assessed 20 patients with unilateral left

hemisphere lesions on their ability to recognize and








discriminate gesture. The patients were divided into four

groups based on the presence of apraxia and anterior versus

posterior lesion location. The apraxic groups did not

differ significantly with respect to the severity of

ideomotor apraxia. However, only the group that was apraxic

with posterior lesions scored significantly worse on gesture

recognition and discrimination. Thus, two forms of

ideomotor apraxia were identified: one based on lesions of

the left inferior parietal lobule which were considered to

damage the visuokinesthetic engrams, and a second based on

lesions anterior to the left supramarginal gyrus which

results in a disconnection of the engrams from the pre-motor

and motor cortex responsible for programming the motor act.

In a further study investigating the dissociations in

apraxic performance, Rothi, Heilman and Watson (1985)

confirmed that LBD apraxic-aphasic patients had poorer

comprehension of gestures than their LBD non-apraxic aphasic

counterparts or normal controls. The importance of the left

parietal region in the production of skilled movement was

also supported by Kolb and Milner (1981), who found that

patients with left hemisphere parietal lesions were

significantly more impaired than patients with right

hemisphere parietal lesions, left hemisphere frontal lesions

or right hemisphere frontal lesions in their ability to copy

meaningless hand gestures.








In addition to distinctions between apraxia secondary

to verbal-motor disconnection, callosal disconnection, and

destruction of the engrams, a fourth subsyndrome was

proposed and identified by Rothi, Mack and Heilman (1986).

They describe two patients with lesions in the inferior

dominant occipital lobe who could gesture and pantomime to

command, could imitate gestures reasonably well and could

gesture appropriately when presented with the tool, but

could not recognize or discriminate gestures and could not

name gestures they imitated. The failure on pantomime

recognition and discrimination could not be explained by

perceptual defects or aphasia. This "pantomime agnosia"

was attributed to a disconnection of the sensory input from

the movement formulae in the inferior parietal lobe of the

left hemisphere. The authors propose that the inferior

visual association cortex is critical for gesture

comprehension, but that the superior visual association

cortex in the left hemisphere is critical for gesture

imitation and/or gesture to tool presentation (p.454).

Watson, Fleet, Rothi and Heilman (1986) described two

cases with damage to the left supplementary motor area (SMA)

in the mesial frontal lobe that resulted in bilateral

ideomotor apraxia, suggesting that the SMA is involved in

motor programming. Watson et al. proposed that the SMA was

required for the translation of the movement representation

to the specific motor program to be executed by the motor






20

cortex. They also suggested that since their second patient

had no evidence of a callosal lesion but was apraxic in both

hands, the left SMA may be critical for praxis in both

hemispheres.

Support for the importance of the SMA in skilled

movement production was provided by regional cerebral blood

flow studies (hereafter referred to as rCBF) which have been

utilized to identify specific regions within the hemispheres

that may be responsible for motor output (Roland, Larsen,

Lassen and Skinhoj, 1980a; Roland, Skinhoj, Lassen and

Larsen, 1980b; Lauritzen, Henriksen and Lassen, 1981).

Normal subjects have been studied at rest and during various

motor tasks. During these motor tasks, increased activity

was observed (as compared to resting levels) in specific

brain regions implicating the activated regions as integral

to the behavior. For example, the execution of intra-

personal motor output (i.e. movements where a body part

moves in relation to itself or another body part such as

finger flexion or finger-to-thumb tapping) resulted in

increased activity in the contralateral "hand region", i.e.

the primary motor and sensory area for the hand, (Lauritzen,

et al. 1981; Roland, et al. 1980a; Roland, et al. 1980b).

With increased task complexity such that sequencing of a

response was required, the SMA, the pre-motor cortex (PMC)

and the inferior frontal region showed increased cortical

activation bilaterally. However, when these movements were








mentally imagined but not executed, Roland and his

colleagues identified increased activation in the SMA but

not the primary motor area, suggesting the role of the SMA

in motor programming (Lauritzen, et al. 1981; Roland, et al.

1980a; Roland, et al. 1980b). These investigators also

observed that when a new motor program was established or a

previously learned program was altered, the convexity pre-

motor regions were activated. Furthermore, there was an

increase in activation in the parietal regions when the

movements took place in extrapersonal space, i.e. with the

environment as the reference instead on the individual's

body, (Roland, et al. 1980b). Shibasaki, Sadato, Lyshkow,

Yonekura, Honda, Nagsmine, Uwazano, Magata, Ikeda, Miyazaki,

Fukuyama, Asato, and Konishi (1993) further supported the

role of the SMA in complex sequential movements, but also

described increases in the primary motor cortex for planning

as well as execution of complex sequential movements.

Halsband, Ito, Tanji and Freund (1993) also stressed the

importance of the left SMA and PMC for reproducing temporal

sequences (as indicated by the ability to reproduce rhythmic

tapping patterns), especially when alternating between left

and right hands. In patients with left hemisphere SMA

and/or PMC damage, both hands were impaired on this task,

whereas patients with right hemisphere damage were impaired

only in the left limb or when each limb was alternated. The

authors suggested that failure of rhythmic tapping following








left hemisphere SMA and/or PMC damage indicated a left

hemisphere dominance for the temporal aspects of motor

programming (Halsband, et al. 1993). Differences in

methodologies likely account for the variance in findings

among these studies. Regardless of these differences, these

studies provide converging evidence to support the

involvement of specific regions previously described in

brain-damaged populations.

In summary, a model for praxis has been developed

consistent with Liepmann's original hypothesis that the left

cerebral hemisphere is dominant for praxis. With respect to

the nature of the praxis representations within the praxis

system, Liepmann (1905) described three components which

formed the basis of a more recent model of praxis (Rothi,

Heilman and Ochipa (1991). First, he identified the

movement formula which he described as being sensory in

nature. He believed these to be visual for the most part,

but acknowledged that they could be represented in other

sensory modalities when those modalities were intrinsic to

the action (Rothi and Heilman, in press). The second

component Liepmann identified was the innervatory pattern,

which transcodes the movement formula into innervation, to

allow for position and movement of the limb. The third

aspect of the system Liepmann outlined were the kinetic

memories, which he believed to be properties of the senso-

motorium and kinesthetic in nature. He believed these








kinetic memories to be stores of highly practiced repeated

movements that could operate independently.

Current models of praxis, such as that of Rothi et al.

(1991) maintain these basic components with some

elaboration. The space-time memories for movement are

considered to be stored in a "three-dimensional, supramodal

code" which must be translated into an innervatory pattern

to then be played out by the motor system (Rothi et al.

1991). Rothi et al. propose that the stores for the

memories, which they refer to as "action lexicons", can be

divided into input and output stores to account for distinct

patterns of receptive or expressive gesture impairment.

Finally, they propose an interface with the semantic system

and the action lexicons, specifically that aspect of

semantics that relates to action. They present this model

in the context of Liepmann's, and those that followed, as

being instantiated in the left hemisphere.

Within the left hemisphere, various regions have been

implicated in ideomotor apraxia, suggesting that the praxis

system is a distributed system which can yield subtypes of

deficit depending upon the localization of the lesion. The

brain regions within the left hemisphere which have been

implicated in apraxia include the inferior parietal region,

frontal regions including the supplementary motor area, the

motor association cortex, the motor cortex, and white matter

regions including the arcuate fasciculus and the corpus

callosum. In addition, lesions which interrupt fibers from








the visual association cortices and from the auditory

cortices within the left hemisphere may also result in

select praxis deficits. Collectively these areas are

assumed to be important for the normal production of skilled

movements. Studies of normal control subjects have

demonstrated increased activation with motor programming and

execution in these regions, lending support to their

importance in skilled movement. Furthermore, studies on

normal control subjects suggest that the left hemisphere is

better suited to program and execute complex sequential

movement.

Right Hemisphere Role in Praxis

Despite his proposal that the left hemisphere was

dominant for praxis, even Liepmann conceded that the right

hemisphere may also have the ability to program and generate

skilled movement (Liepmann, 1905, 1920). He considered the

right hemisphere's ability to generate skilled movement to

be insufficiently developed and therefore dependent on the

left hemisphere. Liepmann believed this right hemisphere

dependence was more pronounced in the absence of tactile-

kinesthetic and visual input, resulting in worse performance

to verbal command or imitation and improved performance with

actual tool use in some cases.

There are several lines of evidence to suggest that the

right hemisphere may have a role in praxis. First, a number

of examples have been cited where lesions that should

produce ideomotor apraxia are nevertheless asymptomatic,








i.e. the so called "negative cases" (Basso, Luzzatti and

Spinnler, 1980; Kertesz and Ferro, 1984; De Renzi,

Faglioni, Lodesani, and Vechi, 1983; Heilman, Rothi, Mack,

Feinberg and Watson, 1986; Graff-Radford, Welsh, and

Godersky, 1987; Faglioni and Scarpa, 1989). Basso et al.

(1980) retrospectively reviewed a continuous series of LBD

patients in an effort to determine the critical lesion

locations for ideomotor apraxia. They reported that

patients with ideomotor apraxia and those without ideomotor

apraxia shared the same lesion locations. Basso et al.

concluded that, apraxiaa is the outcome of widespread

disruption in the left hemisphere and in some callosal

connections or even in the right hemisphere" (Basso et al.

1980, p.124). The authors proposed that the right

hemisphere may be supporting spared praxis functioning in

those patients who fail to demonstrate ideomotor apraxia

subsequent to lesions in the "classical areas" of apraxia

localization within the left hemisphere. Kertesz and Ferro

(1984) reported similar findings. They observed 23 out of

177 patients with large left hemisphere lesions that were

not apraxic. Kertesz and Ferro suggested that the absence

of significant apraxia in these patients was correlated with

significant skull asymmetry, possibly indicating atypical

cerebral organization. They proposed that the absence of

apraxia in some cases may be due to bilateral representation

of praxis function. However, only half of the negative








cases reported by Kertesz and Ferro demonstrated at least

one indication of atypical skull asymmetry. Faglioni and

Scarpa (1989) found no consistent relationship between

hemispheric dominance for praxis and skull asymmetry. In

their study the same percentage of LBD subjects with

atypical cerebral asymmetry as with typical cerebral

asymmetry presented with ideomotor apraxia. This suggested

that the absence of apraxia in some LBD patients was not

related to abnormal or atypical cerebral organization.

These findings, however, do not preclude the possibility of

bilateral representation of praxis, only that it is not

predicted by skull asymmetries.

Right hemisphere pathways for praxis were also proposed

by De Renzi et al. (1983). They found a far greater

incidence of apraxia from parietal lesions than frontal

lesions. This led De Renzi et al. to propose a two-route

model for activation of the right pre-motor cortex similar

to what had been previously proposed by Kliest a half a

century earlier. The first route involves connections from

the left prefrontal cortex to anterior-middle portions of

the corpus callosum, whereas the second route emanates from

the left parietal region crossing at the more posterior

portion of the corpus callosum and then proceeds anteriorly

in the right hemisphere. Thus in those patients for whom

the second route was dominant, it would be possible to cause

a unimanual apraxia from a right hemisphere lesion (De





27

Renzi, et al. 1983). Heilman et al. (1986) described such a

patient. They reported a right-handed 20-year-old patient

with a right superior and inferior parietal lobe lesion who

demonstrated a marked ideomotor apraxia in her left limb.

Her right hand performance was flawless. Most of the

patient's errors were spatial in nature, indicating that the

ability to produce the correct spatial component for gesture

with her left hand had been impaired by a right hemisphere

lesion. She made significantly more errors with her eyes

closed rather than open, indicating that she could improve

performance with visual guidance. Heilman et al. suggested

that their patient was not able to adequately utilize

kinesthetic feedback (which is unilateral) to produce

gestures with the left hand without visual feedback (which

is bilateral). The patient's flawless right hand

performance suggested that each hemisphere was responsible

for programming skilled movements in the contralateral hand.

Graff-Radford et al. (1987) described a young, right-

handed patient with a lesion of the genu and body of the

corpus callosum. The patient's right hand praxis

performance was flawless and left hand performance was

markedly apraxic but improved on imitation. However, Graff-

Radford et al. argued against the theory of posterior

callosal pathways to explain their patient's preserved

imitation ability. Their reasoning was that Liepmann and

Maas's (1907) callosal patient also had sparing of the






28

splenium but failed on gesture imitation. Graff-Radford et

al. suggested that for some patients, the right hemisphere

may be capable of performing visuo-motor tasks (as in

imitation of gesture). Geschwind (1965) also had previously

made a case for the "relative independence" of the right

hemisphere in some patients to program skilled movements in

the absence of input from the left hemisphere in order to

explain how some callosal patients could still imitate

gesture or use tools correctly (Geschwind and Kaplan, 1962).

Even Liepmann's (1900) original case description included

good left arm gesturing to verbal command, imitation and

tool use despite a callosal lesion. This may be evidence

for the right hemisphere's ability to substitute for the

left hemisphere, or it may be evidence for bilateral praxis

representation.

The second line of evidence suggesting a right

hemisphere role in praxis is found in the cases of recovery

from apraxia and "crossed apraxia" which suggest that the

right hemisphere is capable of generating at least certain

aspects of skilled movement (Kramer, Delis and Nakada, 1985;

Rapcsak, Rothi and Heilman, 1987; Rapcsak, Ochipa, Beeson

and Rubens, 1993). Rapcsak, et al. (1993) described a

patient with a massive left CVA who demonstrated that his

right hemisphere had some (albeit flawed) praxis capacity.

Careful praxis testing in this patient revealed good gesture

recognition and discrimination and good tool use. Gesture








to verbal command and imitation was apraxic, with spatial

errors predominating. Rapcsak et al. concluded that because

of the extent of the left hemisphere lesion, the patient's

responses on these tasks had to be supported by the praxis

system of the right hemisphere. They argued that the right

hemisphere is capable of generating skilled movement in

context, whereas the left hemisphere store of

visuokinestetic engrams is required for skilled movement out

of context (such as in gesturing to verbal command). This

possibility is consistent with previous observations that

apraxia is best viewed when tested out of context (De Renzi,

1985).

Further evidence for the right hemisphere's capacity

for praxis is found in two case descriptions of right-handed

individuals who became apraxic from a right hemisphere

lesions (Rapcsak et al. 1987; Kramer et al. 1985). In

neither of these cases could the left hand be tested

secondary to hemiparesis; however, both of them demonstrated

significant ideomotor apraxia in the right hand. This

suggests that for these patients the right hemisphere was

dominant for praxis despite their being right handed.

The third line of evidence that points to right

hemisphere involvement in praxis is that some group studies

have demonstrated that a portion of the RBD subjects score

in the apraxic range (De Renzi, Motti and Nichelli, 1980;

Faglioni and Basso,1985) on tests of skilled movement. While








several studies have been reported which did not find

evidence for apraxia in RBD patients (Liepmann, 1905;

Goodglass and Kaplan, 1963; Lehmkuhl et al. 1983; Duffy and

Duffy, 1989), other studies have found a percentage of the

RBD population who were apraxic (De Renzi, Motti and

Nichelli, 1980; Haaland and Flaherty, 1984; Faglioni and

Basso, 1985). Faglioni and Basso (1985) suggested that the

incidence of apraxia following right hemisphere damage may

be related to the thoroughness of the testing method.

Furthermore, when making group comparisons even a severely

impaired performance of an individual subject may not be

powerful enough to effect the group mean. In Goodglass and

Kaplan's study (1963) there were nine RBD subjects. Of

these nine subjects, six had scores that were in the upper

half of performance accuracy for gesture indicating normal

or near normal praxis. However, two of the subjects had

scores in the middle (suggesting a mild apraxia) and one

subject's scores was near the low end of the range

(suggesting more severe apraxia). Despite this range in

performance, Goodglass and Kaplan did not find statistically

significant differences in praxis scores between the RBD

group and the normal control group. However, there was

clearly one or more cases in RBD group whose praxis

performance would not be considered "normal". Lehmkuhl et

al. (1983) compared 88 LBD patients with 10 RBD patients and

10 normal controls. In spite of finding no significant








differences in praxis performance between the RBD and

control groups, the authors describe one RBD patient who

made errors while performing gestures to verbal command and

to imitation.

De Renzi et al. (1980) tested 100 LBD, 80 RBD and 100

non-brain-damaged controls on a 24 item movement imitation

test. In their study, 16 subjects (20%) in the RBD group

scored in the apraxic range, and of these 16, 4 were

severely apraxic. According to De Renzi et al., the praxis

errors made by the RBD apraxic subgroup did not differ in

nature from the errors made by the LBD group. These

findings were reportedly confirmed in another sample of 110

RBD patients (Faglioni and Basso, 1985). Faglioni and Basso

suggest that the types of gestures sampled may influence the

observation of ideomotor apraxia in RBD individuals. The

more types of movement sampled, the higher the incidence of

apraxia (Faglioni and Basso, p.10). Others have suggested

that specific types of movement are impaired in both LBD and

RBD groups. Dee, Benton, and Van Allen (1970) tested LBD

and RBD patients and found a high incidence of impaired

imitation of meaningless actions in both groups. Kolb and

Milner (1981) found that hemispheric differences for

imitation of hand positions and movements varied based on

which cerebral lobes within the hemisphere were lesioned.

In their study they found left and right frontal ablation

patients equally impaired at copying hand and facial






32

movements. Both frontal lesioned groups were less impaired

than a group of left parietal lesioned patients while right

parietal lesioned patients were unimpaired. Jason (1985)

measured differences between left and right unilateral

excision patients on their ability to generate meaningless

and meaningful finger and hand gestures (i.e. "gesture

fluency"). While the total number of responses did not

differ between groups, Jason found that both left frontal

excision and right frontal excision groups were impaired on

producing "novel" gestures when compared to normal controls.

The final line of evidence that the right hemisphere

may have a role in praxis is that several studies have

identified subtle differences in the performance of skilled

movements among RBD subjects that differ from normal

performance (Kolb and Milner, 1981; Haaland and Flaherty,

1984; Jason, 1985; Heilman, Rothi, Mack, Feinberg and

Watson, 1986; Goodale, Milner, Jakobson and Carey, 1990;

Poizner, Merians, Clark, Macauley, Rothi and Heilman, 1994).

Haaland and Flaherty (1984) compared praxis performance

between 41 LBD and 18 RBD subjects and 25 normal controls in

an effort to identify differences in the types of errors

produced by these groups. All praxis testing was done to

imitation and gestures included transitive, intransitive and

meaningless gestures. Haaland and Flaherty reported that

despite the LBD group being more impaired than the RBD

group, both LBD and RBD groups performed significantly








poorer than the normal controls on all three types of

movements. Furthermore, the LBD and RBD differed from each

other based on the types of errors they produced. While

both groups made more "partial" errors (i.e. errors that

were slowed or less smooth but not a definite error in hand

position, arm position, target location or body-part-as-

object) than their control groups, the LBD group made

significantly more body-part-as-object-1 errors, (where a

part of the body was used as the tool) whereas the RBD made

more body-part-as-object-2 errors (where the hand is in the

correct position but touches the target). The differences

in the quality of the performance between the two brain

damaged groups suggested that each hemisphere may provide

unique information in the production of skilled movement.

Goodale et al. (1990) used a computerized kinematic

analysis to assess RBD subjects on visually guided reaching

tasks (i.e. pointing to a target and line bisection). They

reported that the trajectories for RBD patients reflected a

persistent right-ward visual-motor bias (even though this

was corrected to eventually bisect lines accurately).

Poizner et al. (1994) supported the possibility that RBD

patients may have a deficit in their representation of

external space with the observation that RBD subjects have

difficulty controlling the "plane of motion" of the wrist.

This was the only deficit observed in patients with RBD

using a three-dimensional computerized kinematic analysis to








analyze the temporal and spatial aspects of movement. The

LBD group also demonstrated difficulty in controlling the

wrist plane of motion, but exhibited many other movement

errors as well such as errors in joint synchrony, arm

angles, and spatial and temporal aspects of wrist

trajectories. This would suggest that while the left

hemisphere is critical for skilled movement, the subtle

deficits observed in RBD patients imply there is also a

right hemisphere role in the production of skilled movement.

Harrington and Haaland (1991) investigated right versus

left hemisphere differences in motor sequencing by comparing

RBD and LBD patients on sequences varying in length and

complexity. While the LBD group made significantly more

errors on both repetitive and heterogenous sequences, the

RBD group demonstrated longer delays between the movements

for more complex sequences. Harrington and Haaland

suggested this subtle timing deficit exhibited by the RBD

group occurred on complex sequences because of an increase

in external spatial demands.

In summary, there is evidence to support the idea that

the right hemisphere for some people may be dominant for

praxis. In these cases it may be that there is a reversed

laterality for praxis or bilateral praxis representations.

Studies of cortical activation during skilled movement

support the notion of bilateral praxis representation.

Furthermore, careful measures of subtle aspects of skilled






35

movement suggest that even in individuals with typical left

hemisphere praxis dominance the right hemisphere may have a

role. These latter studies collectively suggest that if

right hemisphere damage results in a deficit in skilled

movement, that deficit may be spatial in nature.

If there is a contribution of the right hemisphere to

praxis processing, the nature of that contribution may be

consistent with functions that have already been ascribed to

the right hemisphere. The descriptions of right hemisphere

errors noted by Haaland and Flaherty provide some support

for the spatial aspect of the RBD groups errors. Assuming

that the right hemisphere's contribution to praxis is based

on typical right hemisphere functions, predictions about the

nature of the right hemisphere contribution may be made.

For example, a spatial influence on praxis processing by the

right hemisphere is consistent with the observation that

lesions in the right hemisphere typically result in spatial

and/or attentional deficits such as neglect (Heilman, Watson

and Valenstein, 1993; Cummings, 1985). Patients with right

hemisphere lesions may fail to bisect lines accurately, may

omit targets on the left side of hemispace, and may neglect

the left side of their own body. Thus, a neglect for the

left side might also interfere with their praxis

performance. Patients with right hemisphere lesions may

also demonstrate spatial disorientation and difficulty

relating body parts to items in the environment (Lezak,






36

1983, Benton and Tranel, 1993). This has been demonstrated

in terms of "dressing apraxia" and also in difficulties

locating items in space. It is possible that right

hemisphere apraxic disturbance may reflect such a visual-

spatial disturbance.

Right-sided lesions have resulted in interference with

temporal relationships and sequences (Lezak, 1983). Defects

in the action-intention system yielding akinesiaa" (failure

of initiation of movement) or "hypokinesia" (delays in

movement initiation) have been attributed to right

hemisphere disease (Heilman et al. 1993). Patients with

these deficits may demonstrate abnormally lengthy delays in

the initiation of movements in response to external stimuli,

or may demonstrate decreased degree of movement. In

addition, right hemisphere lesions may result in motor

impersistence such that they are unable to sustain an act

(Heilman et al. 1993). This would result in the premature

termination of a movement. Finally, motor perseveration may

also result from right hemisphere frontal lesions (Lezak,

1983) yielding inappropriate repeated execution of a motor

response. Sandson and Albert (1984) describe three types of

perseveration that may occur following brain damage. In

particular, the authors attribute continuous perseveration,

i.e., "the inappropriate repetition without interruption, of

a current behavior," (p.730) to a disturbance in motor

output.








Assessment of Apraxia

There are a variety of ways in which ideomotor apraxia

has been evaluated. Differences in test stimuli, test

administration procedures and scoring procedures may account

for some of the differences in findings across studies. For

example, several investigators have found an influence of

the means for eliciting gestures on performance, such that

gesturing by imitation is produced better than gesturing to

auditory command (Alexander et al, 1992; Goodglass and

Kaplan, 1963; De Renzi, 1985; Lehmkuhl et al, 1983).

However, others have found comparable praxis performances

regardless of the means of eliciting the gesture (McDonald,

Tate and Rigby, 1994). Some investigators (De Renzi et al.

1983) have chosen to avoid the possible influence of

impaired auditory comprehension on gesture production by

limiting the mode of input to vision by using an imitation

task. However, using gesture imitation provides the patient

with a model of performance and it has been reported that

providing a model can improve the praxis performance of some

patients (Heilman and Rothi, 1985; Poeck, 1986; Rothi,

Ochipa and Heilman, 1990; Alexander et al, 1992).

Furthermore, others have reported that for some patients,

imitation of gesture may actually be worse than gesturing to

auditory command (Ochipa, Rothi and Heilman, 1994; Mehler,

1987). Thus, the method of elicitation of gesture would








appear to be one factor that needs to be controlled for in

an investigation of left versus right hemisphere praxis

production.

Another factor that has been demonstrated by some

investigators to have an influence on praxis performance is

the type of gesture being produced (i.e. transitive versus

intransitive, static versus sequenced, meaningful versus

non-meaningful). Kimura (1977) reported that LBD patients

were more impaired on producing motor sequences rather than

static postures, however De Renzi et al. (1983) found no

significant differences between sequenced movements and

single movements in identifying errors in praxis performance

for their LBD population. Some investigators have described

greater difficulty imitating non-meaningful gestures among

both brain damaged and normal populations (Pieczuro and

Vignolo, 1967), others reported better performance on

production of non-meaningful (non-representational) gestures

(Alexander et al. 1992), and still others have reported that

there were no significant differences between the production

of meaningful versus meaningless gestures in apraxic

patients (Poeck, 1986; Lehmkuhl, et al. 1983). The

variability of findings among these studies with respect to

the meaningfulness of the targeted gestures indicates the

need to control this aspect of the stimuli in investigations

of apraxia.








Poeck (1986) reported that the distinction between

intransitive and transitive gestures had not been

convincingly demonstrated. However, Maher, Raymer, Foundas,

Rothi and Heilman (1994) described differences in recovery

patterns based on the transitive versus intransitive nature

of the gesture which suggests that these two types of

gesture may be represented differently. Therefore, control

for the possible differences between transitive and

intransitive gestures needs to be considered in

investigations of apraxia.

In addition to the means of eliciting the gestures and

the inherent nature of the gestures, praxis performance

scores may also be influenced by the scoring system and/or

methods used to evaluate performance (Haaland and Flaherty,

1984; De Renzi, 1985; Alexander et al. 1992; Rothi et al.

1988; Poizner, Mack, Verfaellie, Rothi and Heilman, 1990).

The scoring systems used to identify and categorize apraxic

performance have varied considerably across studies, ranging

from behavioral description or binary correct/incorrect

decisions to rating scales and specific error type

identification. The specificity of the scoring criteria may

also have an impact on the identification of limb apraxia in

RBD patients. Rothi et al.(1988) stress that any system

used to describe the errors in praxis performance must

reflect the temporal and spatial nature of limb movement.

McDonald et al. (1994) confirmed the value of analyzing








error type in the investigation of apraxia and it's

underlying mechanisms. Furthermore, it was the error

analysis that led Haaland and Flaherty (1984) to observe

differences between RBD and LBD production of transitive

gestures. The error typing system utilized by Rothi et al.

(1988) allowed for the identification of six error types

that were produced significantly more frequently by left

hemisphere patients than normal controls. McDonald et al.

utilized a system similar to that of Rothi et al. but with

six broad categories and subcategories of error subsumed

under the broader categories. They reported good inter-

judge reliability using this system for detecting the

presence of an error and for the broad type of error.

However, they could not get good reliability in their

efforts to sub-categorize the movement errors. In

conclusion, apraxia has been evaluated in a variety of ways

and there are multiple factors that appear to have a

potential influence on praxis performance. Since there may

be an influence of modality of input, of the type of

gesture, and of the scoring system itself on praxis

performance, these factors need to be controlled in further

investigations of the praxis system.

Questions of This Study

While the left hemisphere clearly has a role to play in

programming the skilled movements of each hand, the role of

the right hemisphere in praxis processing remains unclear.








Therefore, the questions of this study are as follow:

1) Does the right hemisphere contribute to praxis

production? This question will be addressed by determining

if RBD patients produce more errors than normals on a praxis

test known to be sensitive to critical aspects of skilled

movement. If the RBD group does not produce more errors

than normals, we can conclude that impairment of the right

hemisphere does not have an effect on praxis performance.

If the RBD group is different from normals, we can conclude

that the right hemisphere makes some contribution to the

production of skilled movement, and a second question is

raised:

2) What is the nature of the right hemisphere's contribution

to the production of skilled movement? This question will

be addressed by examining the error types produced by the

RBD group compared to the LBD group and non-brain damaged

controls. Three possibilities emerge: A) The contribution

of the right hemisphere to praxis processing is unique and

consistent with what we know about right hemisphere

functioning. B) The contribution of the right hemisphere to

praxis processing is unique but not consistent with other

right hemisphere functions. C) The contribution of the

right hemisphere to praxis processing is not unique but is

consistent with what is contributed by the left hemisphere.

Each of these possibilities has its own set of predictions.

A) The contribution of the right hemisphere to praxis






42

processing is unique and consistent with what we know about

right hemisphere functioning. If this is true then the

nature of the errors produced by the RBD group will reflect

the suspected attributes of right hemisphere involvement.

Specifically, right hemisphere lesions are known to produce

visual/spatial processing problems. Therefore, the praxis

errors may be spatial in nature. External orientation may

deviate to the right as seen in neglect. Movement amplitude

may be greater on the right than on the left. Furthermore,

patients with right parietal lesions may deviate right-ward,

outward and upward when their arm is extended. Hypokinesia

has also been observed subsequent to right hemisphere

lesions. The errors in the RBD may reflect hypokinesia such

that there will be a significant reduction in the amplitude

of the movement, and the movements may be simplified or less

elaborate. Hypokinesia may also be reflected in a reduction

in the cycles of the movement. RBD may result in

impersistence which will also yield a reduction in the

cycles of the movement. Conversely, motor perseveration may

also result from right hemisphere damage. In the case of

praxis, this would be reflected in continuing to produce the

same movement (in reiterative cycles) inappropriately.

B) The contribution of the right hemisphere is unique

but not consistent with other right hemisphere functions.

In this case, there will be no systematic visual/spatial,

hypokinetic or perseverative pattern to the praxis errors.






43

Rather they may reflect behaviors that have not been

previously described with right or left hemisphere lesions.

C) The contribution is not unique but is consistent

with what is contributed by the left hemisphere. In this

case there will be no observable differences between the

nature of the praxis errors produced by the LBD and RBD

groups, though the errors of both groups will differ from

those of the non-brain-damaged controls.














CHAPTER 2
METHODS


The purpose of this study was to determine if the right

hemisphere contributed to the production of skilled

movement, and if so, to describe what element or elements of

movement it may contribute. Using a scoring system that

assesses temporal and spatial aspects of praxis (Rothi et

al, 1988), the performance of skilled acts by right handed

patients with right hemisphere lesions was compared to the

performance of skilled acts by those with left hemisphere

lesions as well as non-brain-damaged controls. Based on the

various factors that can influence apraxia as reviewed in

Chapter 1, the procedures of this study attempted to control

for variables in type of gesture, modality of input,

evaluation methods, and nature of the praxis scoring system.

Since there may be an influence of the type of gesture on

praxis performance, for the purposes of this study gesture

performance was restricted to transitive gestures.

Furthermore, because of the possible influence of modality

of input on praxis performance, in this study the input

modality was limited to auditory command only so as to have

the greatest chance of observing errors. Subjects were








screened to include only those that could follow single

commands. Finally, a scoring system which has been

demonstrated to be sensitive to the errors produced by

patients with ideomotor apraxia (Rothi et al., 1988) was

utilized. However, since it has been demonstrated that

those scoring praxis may have difficulty assessing the

subtler aspects of movement reliably, the study was

structured such that the raters focused on only one aspect

of the gesture at a time.

The data for this study consisted of analyses performed

on archival videotapes from the University of Florida and

Gainesville VA Medical Center Apraxia Laboratories of Dr.

Kenneth M. Heilman and Dr. Leslie Gonzalez Rothi.

Procedures

Right handed individuals with left hemisphere brain

damage (LBD) and right hemisphere brain damage (RBD) and

non-brain-damaged age and education matched, right handed

control subjects were videotaped performing various gestures

to auditory command. These videotaped performances were the

source of the data in this study. Before they could be

analyzed and scored as to the accuracy of the gesture

production, the individual gestures were edited and eight

gestures per subject for each of six subtests were copied

onto master tapes. The subtests were derived from an error-

typing system for apraxia reported by Rothi et al. (1988).

These tapes were then scored by two raters previously








unfamiliar with the present topic who had been trained to

identify correct performance for the gestures. The results

of that scoring provided the data for comparing the

performance of the RBD group to that of the LBD group and

that portion of the normal group whose praxis performance

was recorded while they used their right hand. In addition,

the LBD group performance was compared to that portion of

the control group whose praxis performance was recorded

while they used their left hand (LNC). The six subtests

were summed to yield a total apraxia score, and group

differences were investigated on this total apraxia score

and on the individual subtests. When group differences were

identified, these were investigated further by analyzing

which groups accounted for these differences and by

comparing the error patterns of the significant groups. A

more specific review of these procedures follows below.



Subjects

The subjects for this study were 11 left-hemisphere

brain-damaged patients, 11 right-hemisphere brain-damaged

patients, 11 non-brain-damaged controls using their left

hand and 11 non-brain-damaged controls using their right

hand hereafter referred to as LBD, RBD, LNC and RNC

respectively. All subjects were mono-lingual, English

speaking and right-handed by their own report or the report

of the next of kin. The subjects in the LBD and RBD groups






47

all presented with unilateral, single event cortical lesions

as confirmed by Computerized Axial Tomography or Magnetic

Resonance Imaging (with the exception of one LBD subject who

had EEG but did not have CT or MRI data). Patients with

small subcortical lesions (e.g. lacunes less than 1 CM3) or

periventricular white matter changes consistent with

ischemic demyelination were not excluded. Patient lesion

data are presented in Appendix D. All subjects were free of

seizure disorders, had no psychiatric or prior neurological

history, and were not active alcohol abusers. Each subject

was able to follow single component, auditorily presented

pointing commands as tested informally by the examiner prior

to data collection. Each subject gave consent to

participate in the study. The subject groups were matched

for age, education and time from lesion onset to praxis

testing. Descriptive data on the subjects in this study are

presented in Tables 2-1, 2-2 and 2-3.

Praxis Testing

Each subject was tested individually in a quiet room by

one of five possible examiners. The subjects were asked to

demonstrate the pretended use of 25 common tools, objects or

instruments, using the hand named by the examiner,

ipsilateral to the lesion side. In order to control for

potential variations in task demands for these different

types of items, only the gestures of tool use were used for

the current investigation (see appendix A for a complete









listing of the tools used in this study). There were four

training trials during which the subjects were encouraged to

pretend that they were actually holding the tool in their

hand. They were discouraged from pretending that their hand


Table 2-1:


Subject Descriptive Information: Age


Subject Number
01
02
03
04
05
06
07
08
09
10
11


Mean
Standard Deviation
Median
Range


LBD
45
55
61
51
65
46
63
69
67
66
65


59.36
8.65
63
45-69


Group
LNC
70
60
71
65
47
74
64
76
65
72
70

66.72
8.08
70
47-76


RBD
71
53
66
66
59
66
65
56
68
61
64


63.18
5.38
65
53-71


RNC
60
62
58
65
65
65
66
66
63
73
71


65.45
4.03
65
58-73


Table 2-2: Subject Descriptive Information: Education

Group

Subject Number LBD LNC RBD RNC
01 12 12 15 14
02 12 12 12 12
03 09 12 15 12
04 16 08 13 13
05 13 16 14 13
06 12 06 14 11
07 10 10 08 08
08 14 12 05 14
09 06 18 16 18
10 12 13 12 12
11 08 16 12 16

Mean 11.27 12.27 11.81 13.00
Standard Deviation 2.83 8.08 3.51 4.03
Median 12 12 12 13
Range 6-16 6-18 5-16 8-18








Table 2-3: Subject Descriptive Information: Time post
lesion onset (months) to testing
Group
Subject Number LBD RBD
01 40 60
02 168 65
03 108 23
04 1 9
05 3 16
06 7 11
07 .5 .5
08 48 .5
09 19 148
10 1 .75
11 1 7

Group mean 30.04 30.97
Standard Deviation 54.7 44.95
Median 7 11
Range .5-168 .5-148


was a tool. The first time a subject used his/her hand as

the tool or object, they were corrected by the examiner and

asked to repeat that item pretending that they had the

actual tool or object in their hand. All of the gestures

were videotaped to be reformatted and organized by subtest

for scoring at a later date. If the subject performed a

gesture at an angle that was not easily seen by the camera,

the subject was asked to orient toward the camera and repeat

the gesture, however these instructions varied with the

different examiners.

Basis of Organizing Subject Gesture Production Data for
Subsequent Scoring

Six critical elements of correct gesture production

were derived from the possible praxis error types identified

and described by Rothi et al. (1988), and Greenwald, Rothi,








Maher, Chatterjee, Ochipa and Heilman (1992,a). These six

critical elements formed the basis for the subsequent

reformatting of each subject's recorded productions into six

subtests for later scoring by the raters.

Praxis is a dynamic, nonverbal system that takes place

in a dynamic time/space continuum and by its very nature

difficult to describe verbally. However, every attempt was

made to capture the critical elements of gesture and

variations of those elements for each gesture in a written

description. Two independent experts in ideomotor apraxia

reviewed these critical elements for accuracy. Correct

performance criteria for the gestures selected to assess

each of these identified critical gesture elements are

listed in Appendix B. The six critical gesture elements

(reviewed below in the two headings of tool/object

representations and movement execution) included in this

study were as follows:

Tool/object representation. Successful performance of

a pantomime requires the individual to respect the existence

of an imagined tool and/or object. To do so, he/she must

show spatial accommodation for these absent but imagined

items. Therefore, failure to do so may result in the

following potential errors:

Internal Configuration (IC): When pantomiming, the

fingers/hand/arm must maintain a specific spatial relation

to one another to reflect the presence of the imagined tool.








The internal configuration error refers to any abnormality

of the required forelimb posture. One form of an internal

configuration error has previously been described as a "Body

Part as Tool" error (BPT)(Greenwald et al. 1992a). In this

case the patient uses his/her limb or part of a limb as the

tool itself rather than positioning their forelimb as if

they were holding and using that tool (e.g. the patient uses

his fingers as the blades of the scissors rather than

pretending to hold the scissors). This error type was

described by Goodglass and Kaplan as a Body Part as Object

error. For the purposes of this study, BPT was scored as

incorrect internal configuration.

External Configuration (EC): When pantomiming a

gesture, the fingers, hand, arm and the imagined tool must

be in a specific relationship to the imagined object

receiving the action. External configuration errors refer to

a deficit in orienting the action to the correct part of

space. This may be manifested in a failure to reflect the

appropriate distance necessary to accommodate the imagined

tool to the body or to an imagined object, or may be an

inconsistent orientation. Concretization of the object

receiving the action (CO) (Greenwald et al. 1992a) is a

specialized form of EC error. In this case, instead of

pantomiming tool use on an imaginary object, the patient

pretends to use an imagined tool against a real object such

as a body part or a structure in the room (e.g. the patient








picks up a bit of the hospital gown to cut rather than

imagining the piece of paper). For the purposes of this

study, CO errors were categorized as EC errors.

Movement execution. In addition to tool/object

representations, successful performance of a pantomime

requires the individual to generate a characteristic

movement in peri-personal (near the body) or extra-personal

(away from the body) space. To be correct, this movement

must maintain the accurate plane of trajectory, use the

correct body parts to produce the movement, and be executed

to a sufficient but not excessive degree. Therefore,

failure to do so may result in the following potential

errors:

Movement (M): When acting upon an object with a tool,

a movement characteristic of the action and necessary to

accomplishing the goal is required. Any disturbance in the

characteristic movement may reflect a movement error. This

includes the movement being produced by activity of the

incorrect joints, or using the correct joints but in the

wrong type of action, incoordination of two or more joints

or impaired reciprocal joint action.

Amplitude (A): Each gesture has an acceptable range of

motion size. Errors in the characteristic size of the

pantomime are amplitude errors.

Sequencing (S): There is a characteristic sequence of

the gesture and any perturbation of this characteristic






53

sequence including additions, deletions or transpositions of

movement elements reflects a sequencing error.

Occurrence (0): Certain movements require only one

cycle of the movement, whereas other movements require

multiple cycles or oscillations of the movement. For some

movements, repeated cycles need to remain in a fixed

location, whereas for others the subsequent cycles need to

move across space. Occurrence errors refer to the

inappropriate use of repetitive cycles in cases where

singletons are appropriate or vice versa. They can also be

manifest in the failure of repeated cycles that need to be

altered such that the movement transverses space.

Not all possible elements could be viewed on all of the

gesture stimuli. For example, for the gesture "Use a key to

unlock a door," the number of times the turning gesture

should be made is fixed at one occurrence. If the subject

turns the imaginary key more than once it is considered an

occurrence error. However, for other gestures the number of

cycles of the movement is not constrained (e.g. "Use a

fork"). Consequently, the eating gesture would not be a

good item on which to observe an error of occurrence.

Furthermore, it is more difficult to score a gesture across

multiple performance criteria with good reliability than to

focus scoring decisions on a single element of the gesture

at a time (McDonald et al. 1994). Therefore, the raters

scored only a subset of eight gestures for each critical






54

element (see Appendix B) which had been identified as having

a prototypic performance for that element and could be

viewed easily. The criteria used to select specific

gestures to evaluate a given critical element were 1) it was

possible to produce a violation of that element on the

target gesture, and 2) the prototypic response for the

critical element could be accurately described on the target

gesture. The target gestures for each gesture element were

approved by two experts in praxis testing.

Basis for Gesture Data Exclusion

Prior to the scoring of the production of the gesture

by the raters, each gesture was first rated with respect to

its content (i.e. the semantic reference or meaning of the

gesture) by the trainer. If the content was not accurate,

the gesture was scored as a content error and coded by the

trainer as either a related error (R) (i.e. the gesture was

correct for an associated target item), or as an unrelated

error (U) (i.e., the gesture is recognizable as another

gesture but not related to the target). Finally, a

subject's refusal or failure to produce an attempt at a

gesture was scored as a no response (N). Once the trainer

made the decision that the gesture was a content error and

coded the nature of the content error, no further production

score was given. The incidence of content errors is listed

in Table 2-4. Owing to the low incidence of these errors

they were not formally analyzed. For the purposes of the

analyses these responses were counted as incorrect.








Table 2-4 Content Errors by Group
Error Code LBD LNC RBD RNC
Related 1 0 0 1
Unrelated 1 0 0 0
No Response 2 1 0 0


Rater Training

The two raters for this study were graduate students in

the Departments of Clinical and Health Psychology raterr #1)

and Communication Processes and Disorders raterr #2). The

raters were novices to the study of limb apraxia, were

unfamiliar with the subjects in the study and were unaware

of the purpose and hypotheses of the study. They were

simply informed that people's gesture performance varies

greatly, and that some subjects may have everything correct,

others may have some elements correct and some elements not

correct, and still others may have nothing correct.

Kinematic descriptions of the gestures used in this

study were developed prior to the scoring of the videotapes.

These elemental criteria incorporated the type of motion

(i.e., rotary, linear or curvilinear), the plane of motion,

the direction of the motion (flexion or extension, adduction

or abduction), and the quantity (range) of the motion. The

raters were trained in the definition of the kinematic terms

used in the critical element descriptions during the first

training session. This didactic portion of the first

training lasted approximately one hour, and the raters were

given both written descriptions and illustrations of these






56

terms as part of their rater training packet. A copy of the

rater training packet can be found in Appendix C.

Prior to each scoring session the raters were trained

in the observation and scoring of the particular gesture

element being scored in the subtest for that session (see

Appendix C). The training portion of the sessions included

live and videotaped demonstration of the gesture element

performed correctly for the eight target gestures in that

subtest, and examples of the kinds of errors in that gesture

element that may occur. Following these demonstrations, the

raters practiced evaluating and scoring that gesture element

on videotaped performance of subjects not included in the

present study. These scores were reviewed with the raters

by the trainer (LMM). Any gesture which resulted in

discrepancies in scoring were reevaluated until a consensus

between the two raters was achieved. This procedure

continued until 85% or greater inter-judge reliability

between the two raters was achieved over sixteen trials

(i.e. one trial = one gesture). Once reliability was

achieved the scoring criteria were reviewed one last time

and rating collection began with the test trials. The

amount of time spent on the training of the gesture criteria

and the establishment of reliability for each of the six

scoring sessions ranged from one hour (for Occurrence and

Amplitude) to three and a half hours (for Movement). Data

on inter-rater reliability at the end of the training

portion of each scoring session are presented in Table 2.5.








Inter-Rater and Intra-Rater Scoring Reliability

Inter-rater scoring reliability data for the scoring

portion of each of the six scoring sessions are presented in

Table 2-5. These data were obtained on a subset of the

subjects' scores (N=11) selected at random using a random

number table. A different set of 11 subject scores was

selected for each of the scoring sessions, and the raters

were not aware of which subjects were being used for the

reliability measures. Point-to-point reliability for the

raters was calculated to obtain the percentage of agreement.


Table 2-5: % Agreement Inter-rater Scoring Reliability
Gesture Element % reliability at % reliability for
end of training scoring session

Internal 94% 87% (initial)
Configuration 94% (re-established)

Occurrence 100% 98%

External 94% 84%
Configuration

Amplitude 87% 91%

Movement 100% 77%

Sequencing 87% 86%


In addition to these measures of inter-rater scoring

reliability, three subjects for two of the gesture elements

were randomly selected and were scored twice by the raters

as a measure of intra-rater reliability. These data are

presented in Table 2-6.









Table 2-6 % Agreement Intra-rater Reliability
Gesture Element Rater Number Intra-rater Scoring
Reliability

Sequencing 1 100%
2 96%

Movement 1 92%
2 87%




Scoring

Each subject's performance via videotape was scored by

the two raters and the trainer at the same time but

independently. The rating was completed in a quiet room

using a 20 inch viewing monitor and a Sony video player with

slow motion and freeze frame capabilities. The trainer

controlled the tape player for the raters, but the raters

were free to view the video as many times as they needed in

both regular and slow motion to observe and score each

gesture. They were instructed to evaluate each subject's

performance based on the prototypic target performance

established in the training portion of the session. The

only aspect of the gesture the raters were to evaluate was

the gestural element for that particular subtest. The were

instructed to not be distracted by any other gestural

element that may be produced incorrectly. For example, for

the internal configuration subtest, the raters were to

ignore all other errors that were present, such as a

movement error or an amplitude error and focus solely on the








internal configuration of the gesture. The raters were

encouraged to consult their training packets and the printed

scoring criteria as needed. No discussion was allowed with

respect to the subjects' performance and the raters could

not see each other's or the trainer's score sheet. The

trainer intervened when questions arose as to when to begin

scoring a gesture or which gesture to score (in the case of

a subjects providing multiple responses). Once the scoring

for the first gesture element was completed for all 44

subjects, the scoring session was terminated. The scoring

sessions ranged in time from three hours each for the

occurrence, sequencing and amplitude sessions to five hours

each for the external configuration and movement sessions.

The scoring of internal configuration was split over two

sessions as the kinematic training was completed at the

beginning of that session and we were unable to complete all

44 subjects in that session. As mentioned previously,

scoring reliability of at least 85% was re-established in

the second IC scoring session.


Analysis



Once the raters had completed scoring the six subtests

in this study, the scores were tabulated. If there was a

disagreement between the two raters, the trainer's score was

used to break the difference. The videotapes were reviewed








again by the trainer and for each item that was scored as

incorrect an attempt was made to describe each error so that

patterns of performance could be analyzed later. For each

gesture element a classification for the described errors

was developed. This classification system is presented in

Table 2-7. There were occasions when the error could not be

classified because the patient either refused to respond or

made a content error, or because the trainer had disagreed

with both raters and scored the item as correct. In these

instances the error was tallied as "not coded".








Table 2-7: Classification of Errors


Gesture Element


Internal
configuration









Occurrence








External
configuration


Category of Error


1. Body Part used as the tool
2. Absence of space in the hand to
accommodate the tool
3. Tool space in the hand too large for
the tool
4. Tool "held" incorrectly (e.g. saw
held by fingertips)
5. Correct position not maintained over
the course of the movement
6. Not coded

1. Too many repetitions
2. Too few repetitions
3. Repetition produced in the same space
where transversing space is
required (e.g. stays in same space
on the face when shaving with a
razor)
4. Not coded

1. Plane of object not maintained (e.g.
hand breaks through the coronal
plane when erasing the blackboard)
2. "Object" changes location
3. Initial orientation not correct (e.g.
hand oriented to floor for erasing
a blackboard)
4. Not oriented toward self correctly
(e.g. hand not near enough to mouth
for brushing teeth)
5. Failed to leave space for "tool"
between hand and object or self
6. "CO"; uses an object or body part to
represent the object
7. Not coded


(Table 2-7 continued on next page)








Table 2-7 (continued)


Gesture Element


Amplitude






Movement


Sequencing


Category of Error


1. Movement too small
2. Movement too large
3. Amplitude varies when it should
remain stable
4. Not coded


1. Correct joints move but the action is
incorrect
2. Fails to stabilize certain joint(s)
as required to complete the correct
action
3. Joint action missing
4. Partial movement required but instead
done completely
5. Additional, extraneous movements
6. Not coded

1. Correct movements but out of sequence
2. Part of sequence missing
3. Hand not positioned prior to movement
4. Elements of a sequence repeated
5. Not coded













CHAPTER 3
RESULTS


While the literature suggests a role for the left

hemisphere in programming the skilled movements of each

hand, the role of the right hemisphere in praxis processing

remains unclear. This study was conducted in an effort to

determine if the right hemisphere contributed to the

production of skilled movement, and if so, to describe what

element or elements of movement it may contribute. This was

attempted by comparing the performances of RBD subjects with

those of LBD, LNC, and RNC subjects. The critical gestural

elements that were used for the assessment of praxis

performance were derived from a scoring system that assesses

temporal and spatial aspects of praxis (Rothi et al., 1988).

These six critical gestural elements were assessed in six

separate subtests consisting of eight gestures each produced

to an auditory command. Overall praxis performance was

determined by summing the six subtest scores. Two raters

who were previously unfamiliar with limb apraxia or the

subjects in this study were trained on the correct

performance of each of the tool-use gestures. Following

this training the raters scored the archival videotaped








performances of the experimental groups. Inter-rater

reliability for the scoring ranged from 77% to 98% agreement

over the six subtests, with a mean inter-rater scoring

reliability of 87%.



Analysis of Group Characteristics

Initial review of the data using the descriptive

statistics presented in Chapter 2 and plots of normal

probability indicated that the data did not meet the

assumptions of normality required for parametric statistical

procedures. Furthermore, an assessment of ANOVA residuals

for some of the measures suggested violation of the

assumption of homogeneity of variance. Thus, the data in

this study were analyzed by nonparametric measures. All the

analyses in this study were completed using SYSTAT for

Windows, Version 5 (Wilkinson, Hill, Welna and Birkenbeuel,

1992).

The Kruskal-Wallis One-Way Analysis of Variance

nonparametric procedure indicated that the four experimental

groups were not significantly different with respect to age

(H = 5.36, p > .05) or education (H = 1.63, p > .05). The

Mann-Whitney U test, a nonparametric measure for comparing

two independent groups, indicated that the two brain-damaged

groups were not significantly different with respect to the

amount of time between lesion onset and time of testing (U =

60.5, p > .05).






65

Gesture performance was divided into six subtests, each

designed to assess only one critical gestural element

(internal configuration, external configuration, occurrence,

movement, amplitude, and sequencing). Thus, each subtest

score was thought to be a reflection of the degree to which

a subject displayed a deficit of that critical element of

movement. Scores of accuracy on the six subtests were

summed to yield a total apraxia score thought to reflect the

overall degree of praxis impairment displayed by the

subject. Descriptive statistics for total apraxia score and

the six gesture element subtests including means, standard

deviations, medians and ranges for the four experimental

groups are presented in Tables 3-1 through 3-7.


Table 3-1 Number Correct for Total Apraxia Score By Group

LBD LNC RBD RNC

Mean 30 38.72 37.9 38.18
Standard Deviation 5.47 4.88 3.53 5.23
Median 28 39 37 39
Range 23-40 30-46 34-43 29-46




Table 3-2 Number Correct for Internal Configuration Subtest
By Group
LBD LNC RBD RNC

Mean 3.27 5.09 4.81 4.81
Standard Deviation 1.34 2.16 1.94 1.60
Median 4 5 5 5
Range 2-6 2-8 1-7 3-7








Table 3-3 Number Correct for External Configuration Subtest
By Group

LBD LNC RBD RNC

Mean 4.27 6.18 5.45 5.81
Standard Deviation 2.32 1.07 1.80 1.25
Median 5 6 6 6
Range 0-8 4-8 3-8 4-8


Table 3-4 Number Correct for Occurrence Subtest By Group

LBD LNC RBD RNC

Mean 6.90 7.36 7.27 7.09
Standard Deviation 0.83 0.67 0.78 1.04
Median 7 7 7 7
Range 6-8 6-8 6-8 5-8


Table 3-5 Number Correct for Movement Subtest By Group

LBD LNC RBD RNC

Mean 3.09 6.09 6.54 6.00
Standard Deviation 1.30 1.51 1.29 1.94
Median 3 6 6 6
Range 1-6 4-8 5-8 2-8


Table 3-6 Number Correct for Amplitude Subtest By Group

LBD LNC RBD RNC

Mean 6.09 6.63 6.18 7.54
Standard Deviation 2.32 1.07 1.80 1.25
Median 6 7 6 8
Range 3-8 5-8 4-8 6-8


Table 3-7 Number Correct for Sequencing Subtest By Group

LBD LNC RBD RNC

Mean 6.36 7.36 7.63 6.90
Standard Deviation 1.36 0.80 0.50 0.83
Median 6 8 8 7
Range 4-8 6-8 7-8 6-8








Research Questions

Research Question # 1

1) Does the right hemisphere contribute to praxis

production?

While the majority of the literature supports the

theory that the left hemisphere is dominant for praxis,

several studies have provided evidence that limb apraxia may

also result from right hemisphere damage. This question was

addressed initially by determining if the RBD group produced

more errors than the RNC group on the total praxis measure.

Secondly, individual subtest scores of the RBD subjects were

compared to those of the RNC group to determine if there

were differences between the two groups that might have been

masked when the subtest scores were combined. Thus the null

hypotheses for both parts of the question were stated:

Ho: There are no significant differences between the

experimental groups on the total apraxia score.

Ho: There are no significant differences between the

experimental groups on the individual subtests of

performance.

If RBD group did not produce more errors than the RNC group

on either the total apraxia score or the individual

subtests, we would conclude that impairment of the right

hemisphere does not have an effect on praxis performance.

If the RBD group was different from the RNC group, we would

conclude that the right hemisphere may contribute in some








way to the production of skilled movement, and further

questions as to the nature of that contribution would be

raised.

The first part of Question #1 (i.e. group differences

on total apraxia score) was addressed using the Kruskal-

Wallis One-Way Analysis of Variance nonparametric procedure.

The results of this analysis suggested that there was a

significant difference between the four groups on total

apraxia score (H = 13.32, p = .004) and the follow-up

pairwise comparisons using the Mann-Whitney U test are

listed in Table 3-8. Significance levels for the pairwise

comparisons were adjusted to .01 to control for Type I error

owing to the multiple comparisons being made.



Table 3-8 Pairwise Comparisons for Total Apraxia Score

Groups LBD vs LNC RBD vs RNC LBD vs RBD LNC vs RNC
Rank
Sum 79 vs 173 123 vs 129 81 vs 171 130 vs 122

U Stat. 13.5 57.5 15.5 64.5

p-value .001* .842 .002* .791

* = significant difference


The results of this analysis suggest that the difference

found between the four experimental groups can be accounted

for by the LBD group. Their performance was significantly

worse than that of the LNC group and that of the RBD group.








In contrast, the other groups (of particular interest, the

RBD group) did not differ significantly from each other on

the total apraxia score.

The second portion of Question 1 (i.e. were there group

differences on the individual gesture element subtests) was

addressed using the Kruskal-Wallis One-Way Analysis of

Variance nonparametric procedure. The results of these

analyses are presented in Tables 3-9 through 3-14.

Significance levels were adjusted in order to control for

Type I error owing to the number of comparisons being made.

Since apriori it was believed that some of the subtest

comparisons would be more important than others in

indicating group differences (i.e. based on the functions of

the right hemisphere and known effects of right hemisphere

disfunction) the alpha levels required for significance were

weighted differently across the six subtests. It was

predicted that the two most fruitful comparisons would be

for the external configuration and amplitude subtests.

Therefore, the alpha level for these two subtests was set at

.023. The alpha levels for the other four subtests

(occurrence, sequencing, internal configuration and

movement) were each set at .001. Thus, the combined alpha

level required for significance for the six subtests was

.05.

The results of this analysis suggested that there were

significant group differences on the amplitude and movement








subtests. Mann-Whitney U tests were utilized to make

comparisons between the groups on these two subtests to

determine which groups accounted for these differences.

Table 3-9 Kruskal-Wallis ANOVA for Gesture Element Subtests


Subtest H Stat. p-value

Internal Configuration 6.543 .088

External Configuration 6.002 .112

Occurrence 1.838 .607

Amplitude 10.171 .017*

Movement 18.592 .0003*

Sequencing 8.339 .039**


* = significant difference
**= trend

Comparisons were made between the brain-damaged groups and

their respective control groups as well as a comparison of

both brain damaged groups and both control groups. The

results of these pairwise comparisons for the movement and

amplitude subtests are listed in Table 3-10. Significance

levels for the pairwise comparisons were adjusted to .01 to

control for Type I error due to the number of comparisons.

The results of these comparisons suggest that for the

movement subtest the difference between the groups can be

accounted for by the poor performance of the LBD group on

this subtest. The RBD group was not significantly different

from their control group but was different from the LBD

group. For the amplitude subtest, the RBD group did perform








differently than their control group suggesting an

impairment on the amplitude gestural element. Furthermore,

the RBD group was not significantly different from the LBD

group on this subtest.


Table 3-10 Pairwise Comparisons for

Movement

Groups LBD vs LNC RBD vs RNC
Rank
Sum 73 vs 179 133 vs. 119

U Stat. 7.5 67.5

p-value .0004* .636


Movement and Amplitude


LBD vs RBD

70 vs. 182

4.5

.0001*


LNC vs RNC

125 vs 128

59.0

.919


LBD vs LNC

114 vs 139

48.0

.394


Amplitude

RBD vs RNC

86 vs 167

20.0

.005*


LBD vs RBD

126 vs 127

60.0

.972


LNC vs RNC

94 vs 158

28.5

.025**


* = significant difference
** = trend



Research Question # 2

2) What is the nature of the right hemisphere's contribution

to the production of skilled movement?

Because the RBD group did demonstrate an impairment on

one of the experimental measures, follow-up was needed to

see if there was anything specific in the performance of the


Groups
Rank
Sum

U Stat.

p-value





72

RBD group that would help distinguish it from the LBD group.

To address this issue, a comparison of error types was made

between the RBD and LBD groups. Three possibilities were

recognized: A) The contribution of the right hemisphere to

praxis processing is unique and consistent with what we know

about right hemisphere functioning. In this case a

preponderance of spatial errors was predicted in the form of

external configuration and/or amplitude errors consistent

with impaired right hemisphere visual/spatial processing.

Reduction in amplitude of the movement or in the number of

cycles of the movement consistent with hypokinesia or motor

impersistence, which have been observed subsequent to right

hemisphere lesions was also predicted. Conversely, motor

perseveration might have yielded occurrence errors in the

form of inappropriate reiterative cycles.

B) The second possibility was that the contribution of the

right hemisphere to praxis processing was unique but not

consistent with other right hemisphere functions, resulting

in no systematic visual/spatial, hypokinetic or

perseverative pattern to the praxis errors.

C) The third possibility was that the contribution of the

right hemisphere to praxis processing was not unique but was

consistent with what was contributed by the left hemisphere,

in which case there would have been no observable

differences between the nature of the praxis errors produced

by the LBD and RBD groups, though the performance of both

groups would have differed from those of the non-brain-

damaged controls.






73

Differences between the groups for types of errors were

analyzed using the error categories listed in Table 2-7.

The frequencies of each error type were tallied by group for

each gesture element subtest. These data are presented in

Tables 3-11 through 3-16. Data on the error performances of

each control group are also presented for comparison. Chi

square analysis could not be performed on these data due to

the number of cells with values less than 5. To identify

differences in error patterns of the RBD and LBD groups, pie

charts depicting the proportion of each error type for the

two subtests that had yielded significant results (amplitude

and movement) were created. These data are presented in

Figures 3-1 and 3-2. While they must be compared with

caution because the total number of errors differs between

the two groups, a relatively comparable performance between

the two groups with respect to amplitude error types was

observed. Conversely, a difference in error pattern between

the two groups can be identified on the movement subtest in

that the majority of the error occurs on incorrect joint

selection, activation or stabilization for the LBD group,

while the RBD group made relatively few of these errors.








Table 3-11 Total of Error Types on Internal Configuration

Error Category LBD RBD LNC RNC

1. Body Part used as the tool 11 9 5 4
2. Absence of space in hand 23 21 19 14
to accommodate the tool
3. Tool space too large 3 0 0 0
4. Tool "held" incorrectly 11 3 4 13
5. Position not maintained 1 1 2 0
during movement
6. Not coded 2 0 1 2


Table 3-12 Total of Error Types on Occurrence

Error Category LBD RBD LNC RNC

1. Too many repetitions 6 2 4 4
2. Too few repetitions 1 0 0 0
3. Repetition in the same 3 6 1 6
space when transversing
space is required
4. Not coded 2 0 2 0



Table 3-13 Total of Error Types on External Configuration

Error Category LBD RBD LNC RNC

1. Plane not maintained 16 11 5 9
2. "Object" changes location 8 4 3 5
3. Initial orientation wrong 5 4 3 2
4. Self-orientation incorrect 7 3 4 4
5. Absence of space for "tool" 2 3 3 4
6. CO 7 2 0 1
7. Not coded 0 0 0 0

Table 3-14 Total of Error Types on Amplitude

Error Category LBD RBD LNC RNC

1. Movement too small 7 3 3 3
2. Movement too large 14 17 11 2
3. Amplitude varies 0 1 0 0
inappropriately
4. Not coded 1 0 1 0








Table 3-15 Total of Error Types on Movement

Error Category LBD RBD LNC RNC

1. Correct joints but 11 2 1 3
incorrect action
2. Fails to stabilize 11 3 8 6
certain joint(s)
3. Joint action missing 26 4 11 7
4. Partial movement required 4 1 1 2
but instead done completely
5. Additional, extraneous movement 0 0 0 2
6. Not coded 5 6 4 3



Table 3-16 Total of Error Types on Sequencing

Error Category LBD RBD LNC RNC

1. Movements out of sequence 3 2 0 2
2. Part of sequence missing 6 1 2 5
3. Hand not positioned 7 1 0 6
prior to movement
4. Elements of the sequence 1 0 2 1
repeated inappropriately
5. Not coded 0 0 2 0








Figure 3-1:Percent of Total Error for Error Classifications
on the Amplitude Subtest for the RBD and LBD Groups


LBD GROUP


RBD GROUP


TOOSMALL


NOCODE


TOOLARGE


TOOLARGE


TOOLARGE = Movement larger
than appropriate
TOOSMALL = Movement smaller
than appropriate
Variable = Movement varies in
amplitude when it should not
Nocode = Error not coded


TOOSMALL



VARIABLE


LBD
64%

32%

0%


RBD
81%

13%

5%








Figure 3-2:Percent of Total Error for Error Classifications
on the Movement Subtest for the RBD and LBD Groups


LBD GROUP


ACTMISSING


RBD GROUP


EXTREME


ACTMISSING


NOCODE


WRONGACT


NOCODE


NOTSTABLE


LBD


RBD


WRONGACT = 1. Correct joints 19% 12%
are activated but with the
incorrect action
NOTSTABLE = 2. Fails to stabilize 19% 19%
joints in the action
ACTMISSING = 3. A joint action is 46% 25%
missing from the movement
EXTREME = 4. A partial movement 7% 6%
was required but it was done
completely.
ADD = 5. Additional movements 0% 0%
NOCODE = 6. Not coded 8% 37%0
O Not coded items for RBD group were all due to trainer
disagreement with raters.


rREME









Extraneous Variables



In the course of scoring the videotapes it became

apparent that the five examiners did not use the same verbal

commands to elicit the gestures from the subjects.

Specifically, three of the examiners tended to provide more

bits of information that reduced the degrees of freedom in

the gesture performance. For example, the verbal command to

produce the gesture for key was given sometimes as "Show me

how you would use a key" and at other times as "Show me how

you would use a key to unlock a door in front of you." (For

a list of both command forms see Appendix A.) Both of these

methods have implications for performance and might have

been driven by the biases of the examiners. The motivation

for the embellished commands possibly was based on the need

to be certain that the gesture was produced in a direction

and orientation that maximized the visibility for scoring

later. Alternatively, it may have been in response to the

subjects themselves since the brain-damaged populations

tended to receive the more embellished commands.

To investigate the impression that the brain-damaged

subjects received the embellished commands, the tapes were

reviewed again by the trainer. Each tool that was used in

the study was examined for the amount of information

provided in eliciting the gesture. Each tool was coded only

once because the same clip was used whenever a subtest






79

called for that gesture. A point was given for each bit of

information on each command to yield an "index of

embellishment". For example, the command "Show how you

would use a key" would yield one point, since the only bit

of information is provided by the tool name. Alternatively,

the command "Show how you would use a key to unlock a door

out in front of you" would receive a score of four; one for

the tool name, one for the verb action, one for the

recipient of the action and one for the location of the

action. In this way, the degree of additional information

provided could be assessed to determine if there was a

systematic bias in the data. The results of that tally are

listed in Table 3-17 as total bits of information.

Table 3-17 Index of Embellishment by Group
Group
Subject # LBD LNC RBD RNC
01 58 26 44 30
02 29 33 46 34
03 45 32 55 29
04 28 31 53 28
05 41 44 48 26
06 27 41 26 34
07 29 45 47 23
08 33 41 26 24
09 36 43 46 31
10 33 54 32 32
11 36 45 50 49

Group Mean 35.9 39.54 43.0 30.9

These data suggest that the LBD and LNC groups received

approximately the same degree of information, however the

RNC group received considerably less information than the

RBD.








This difference in the amount of information some of

the groups received in eliciting the gestures may have

introduced variability for the identification of error

patterns between the two brain-damaged groups. To

illustrate, Table 3-18 provides the tally of which groups

received information with respect to the location of the

gesture on the EC subtest.



Table 3-18: Number of Location Cues Provided By Group
LBD LNC RBD RNC
# Location Cues Given 17 34 41 6

# EC Gestures scored 41 20 28 22
as incorrect

# EC Gestures scored 5 8 15 4
as incorrect when
location provided

# EC gestures scored 36 12 13 18
as incorrect when
location not provided


As a result of this variability in how the gestures

were obtained, the follow-up analyses planned for exploring

the nature of the RBD groups errors could not be completed.

Hemi-spatial deficits could not explored because the

comparison groups were not given the same location

information as the RBD group. Additionally, there was great

variability with respect to the amount of time each

individual subject was given to perform the gesture before

they were asked to produce the next item. Therefore,

patterns of motor impersistence or motor perseveration could

not be explored.








Summary

The purpose of this study was to determine if the right

hemisphere was involved in the production of skilled

movement by comparing gesture production of RBD subjects

with those of RNC and LBD subjects. The results of this

study suggest that the RBD group was not different from the

RNC group with respect to total apraxia score or on most of

the gesture element subtests. However, the RBD group did

differ from its control group (RNC) on the amplitude

subtest, but did not differ from the LBD group on this

measure. The movement subtest distinguished the LBD group

from the RBD and LNC groups. The two normal control groups

did not differ significantly on any of the measures in this

study with the exception of a trend for differences on the

amplitude measure. Descriptive error pattern analysis

revealed differences between LBD and RBD groups for the

movement subtest but not for the amplitude subtest. More

explicit hypotheses about the RBD performance and right

hemisphere processing were not explored due to methodologic

issues in data acquisition.












CHAPTER 4
DISCUSSION


The dominant role of the left hemisphere in the

production of learned skilled movement (i.e. praxis) has

been emphasized in the literature since Liepmann first made

the proposal in the early 1900's. However, there have been

recent suggestions that the right hemisphere may also play a

role in praxis. The purpose of this study was to determine

if the right hemisphere is involved in the production of

skilled movement by comparing right hemisphere brain-damaged

subjects to non-brain-damaged controls and to left

hemisphere brain-damaged subjects on a gesture to auditory

command task. Eleven subjects with right hemisphere

unilateral cortical lesions, eleven subjects with left

hemisphere unilateral cortical lesions, eleven non-brain-

damaged controls using their right hand and eleven non-

brain-damaged controls using their left hand comprised the

subjects for this study. Videotapes of these subjects

producing gestures to auditory command were made by five

different examiners. These tapes were edited to select

eight gestures for each of six subtests of praxis

performance. Two raters who were unfamiliar with the

disorder of apraxia or the hypotheses of this study were








trained to score the gesture performances on these six

subtests. Each subtest focused on the assessment of one

critical gestural element believed to be important for

accurate gesture production.

Scoring precision was encouraged by requiring that the

two raters use the criteria outlined for them on correct

performance for each critical element. They scored the data

independently and achieved good inter-judge and intra-judge

reliability. The scores of the individual gesture element

subtests were summed to yield a total apraxia score. This

chapter will discuss the current study with respect to: 1)

conclusions regarding the research questions posed in this

study, 2) methodological issues, 3) clinical implications,

4) implications for future research.



Research Questions



The primary question of interest for this study was:

Does the right hemisphere contribute to the production of

skilled movement? The major finding of this study was that

patients with right hemisphere brain damage did not display

performance on total apraxia score that was significantly

different than normal control subjects using their right

hands. This suggests that overall the right hemisphere

brain-damaged group did not demonstrate ideomotor apraxia

and yields the conclusion that the right hemisphere does not








have a major role in the motor programming of skilled

movement. In contrast, and not unexpectedly, the left

hemisphere brain-damaged group did demonstrate ideomotor

apraxia as evidenced by a significant difference between

their performance and that of the normal control group using

their left hands for the total apraxia score.

These findings contradict those of Haaland and Flaherty

(1984) who reported that when they used an error typing

system that was sensitive to the qualitative aspects of

movement they were able to identify impaired performance in

a group of right-handed right hemisphere brain-damaged

patients. They further reported that the errors made on

transitive gestures were qualitatively different than those

errors produced by subjects with left hemisphere brain

damage. However, the findings in the current study are

consistent with those described by Liepmann and others

(Liepmann, 1905; Goodglass and Kaplan, 1963; De Renzi et al.

1968; Lehmkuhl et al. 1983; Duffy and Duffy, 1989; Alexander

et al. 1992). The present investigation differs from the

Haaland and Flaherty study in several ways. First, the

raters in this study were unfamiliar with the hypotheses and

with the historical literature, and as such were relatively

unbiased with respect to their expectations for the

subjects' performances. Haaland and Flaherty did not use

unbiased scorers which may have contaminated their results.

Second, the scoring in the present study was structured such








that the individual elements of skilled movement to be

evaluated were scored in isolation, allowing the raters to

focus their attention on only one dimension of the movement.

This hopefully resulted in more precise judgements than when

all dimensions of movement are scored at the same time as in

Haaland and Flaherty's study.

Another way in which these two studies differed was in

how the gestures were categorized. The critical gesture

elements in the present study were devised to be mutually

exclusive such each type of error was independent (i.e.

making one did not prevent the subject from being assessed

on the other gesture elements). For example, the categories

in the current study considered body-part-as-tool errors as

a form of internal configuration error because if a subject

makes a body-part-as-tool error they cannot possibly make a

hand position error (the way they are described by Haaland

and Flaherty). However, in the Haaland and Flaherty study,

they scored body-part-as-tool (they use the term "object")

and hand position errors in two error categories.

Therefore, there was not an equal chance of making each type

of error (i.e., the variables were not independent). This

may have contributed to the differences in the incidence of

these errors between the two studies, since in the present

study those aspects of skilled movement that were believed

to be independent were evaluated. Other differences between

the current study and the Haaland and Flaherty study will be








addressed in the subsequent section. As stated in the

results, combining all six subtests may have masked group

differences on the individual measures of internal

configuration, external configuration, occurrence,

sequencing, amplitude and movement. Therefore, the four

experimental groups were also compared on these individual

gesture element subtests. Based on what previously had been

described in right hemisphere brain damage (Haaland and

Flaherty, 1984; Benton and Tranel, 1993) it was predicted

that the right hemisphere brain-damaged group would more

likely demonstrate difficulty with the external

configuration subtest and the amplitude subtest. However,

this prediction was only partially upheld. The single

dimension on which the right hemisphere brain-damaged group

performed differently than normals was amplitude. In fact,

the group differences on the amplitude subtest were

accounted for by the exceptionally good performance by the

normal controls using their right hand and the poor

performance of both brain-damaged groups. The two brain-

damaged groups were equally impaired on this measure, though

the left hemisphere brain-damaged group was not

statistically different from their control group. It is

perhaps not surprising that the normal control group using

their dominant hand produced gestures somewhat better

(although not enough to be statistically significant) than

the normal control group using their non-dominant hand.






87

This seemed to be a factor only for the amplitude dimension.

However, hand dominance cannot explain the poor performance

of the right hemisphere brain-damaged group since they were

using their dominant right hand.

In addition to amplitude, the only other measure to

yield significant group differences was movement. In

contrast to amplitude, follow-up comparisons indicated that

the group differences on the movement subtest could be

accounted for by the poor performance of the left hemisphere

brain-damaged group alone. The other elements that were

believed to be critical to the production of skilled

movement failed to achieve significance, even for the left

hemisphere brain-damaged group. In particular, external

configuration failed to distinguish either brain-damaged

group from their normal control groups. Possible

explanations for this negative finding will be discussed in

the subsequent section on methodologic issues.

The finding that there was one dimension that

distinguished the right hemisphere brain-damaged group

performance from that of their normal controls justified the

second question in this study: #2. What is the nature of

the right hemisphere's contribution to the production of

skilled movement? A descriptive error typing analysis was

made of the amplitude errors for both brain-damaged groups,

however no specific differences in the profile of the errors

was found between the two groups. This may mean that the






88

information provided by the two hemispheres with respect to

the size (amplitude) of the movement may be the same.

Conversely, the two brain-damaged groups may fail for

different reasons. As was stated above, the normal controls

using their left hand were not significantly better in their

performance on the amplitude dimension than the left

hemisphere brain-damaged group, hinting that possibly there

may be an effect of the use of the nondominant hand. Use of

the nondominant hand for writing (one of the items on this

subtest) has been demonstrated to result in proximalizing

the action of the movement, yielding greater elbow

displacement which visually may appear as the size of the

movement(Mack, Rothi and Heilman, 1993). However this would

explain only errors of increased amplitude, and both groups

also made errors of decreased amplitude.

The error type analysis did reveal a marked difference

between the two brain-damaged groups for movement in that

compared to the right hemisphere brain-damaged group, the

left hemisphere brain-damaged group made more joint action

errors, such as failing to use a particular joint or using

the joint but in the incorrect action. The right hemisphere

brain-damaged group made relatively few errors of this type

confirming that there were differences in performance

between the two groups along some aspects of the gesture.

The difficulty of the left hemisphere brain-damaged group

with joint activation and control is consistent with what

has been reported using kinematic analyses of gestures in








left brain-damaged, apraxic subjects (Poizner et al. 1990;

Poizner, Clark, Merians, Macauley, Rothi and Heilman, 1995).

These studies have demonstrated in an objective measurement

that left hemisphere brain-damaged subjects have marked

difficulty with joint coordination and with translating the

spatial plan for the movement into the angular movements of

the joints.

Kinematic analyses may provide some insight as to the

right hemisphere brain-damaged group performance in this

study as well. Poizner and his colleagues have demonstrated

that the only dimension they observed to be impaired in

right hemisphere brain-damaged subjects in their study was

in maintaining an accurate plane of wrist motion in space

(Poizner, Merians, Clark, Macauley, Rothi and Heilman,

1994). They proposed that this difficulty may be related to

right hemisphere deficits in the representation of

extrapersonal space, since maintaining an accurate motion

requires a determination of the pretended object in external

space. These same researchers have also described deficits

in the spatio-temporal features of the movement in left

hemisphere brain-damaged apraxic subjects (Clarke, Merians,

Kothari, Poizner, Macauley, Rothi and Heilman, 1994),

attributing the left hemisphere brain-damaged subjects'

performance more to a deficit in the movement plan. This

hypothesis is consistent with a case report of a patient

with severe ideomotor apraxia subsequent to left hemisphere

brain damage who demonstrated preserved knowledge regarding








whether the gestures were produced in peri-personal versus

extrapersonal space. This suggested that knowledge regarding

whether a particular gesture is produced in peri-personal

versus extrapersonal space is not intrinsic to the movement

plan (Maher, Raymer, Rothi and Heilman, 1993). The data in

the present study neither confirm nor refute these claims.

The brain-damaged groups in the present study did

demonstrate a deficit in the amplitude of the movement. The

left hemisphere brain-damaged group clearly demonstrated

difficulty with joint coordination and the plan of the

movement, which may have contributed to the amplitude

impairment. Furthermore, it is plausible that difficulty

with spatial relations in extrapersonal space in the right

hemisphere brain-damaged group may have yielded their errors

in amplitude. Thus, the impaired performance of the two

groups may have different causes. However, based simply on

the results of this study, there is not enough evidence to

state that the nature of the amplitude deficit for the two

brain-damaged groups is different.

Methodologic Issues/Study Limitations

Some of the predictions in this study with respect to

right hemisphere functioning were not upheld, and others

were unable to be assessed due to methodologic constraints.

A deficit in the number of cycles for movement (occurrence)

was not observed in either brain-damaged group. The lack of

occurrence error in the left hemisphere brain-damaged group

is consistent with what Rothi et al.(1988) reported in their








study of left-hemisphere subjects and apraxic errors.

However, in the present study it was predicted that the

right hemisphere brain-damaged group may demonstrate motor

impersistence resulting in a decrease in the number of

repetitive cycles, or motor perseveration resulting in an

increase of repetition in cycles when inappropriate.

Neither of these deficits was observed. The time allowed to

respond to the auditory command varied greatly among the

examiners making it impossible to reliably measure motor

perseveration or motor impersistence. Furthermore, only one

gesture in this subtest required a single occilation, making

the observation of perseverative errors in the form of

repeated cycles very low. There was a tendency for some of

the right hemisphere subjects to demonstrate what Sandson

and Albert (1984) described as "continuous perseveration" in

that they continued producing the gesture until the examiner

gave the next gesture command. The normal controls were

stopped after only two or three cycles of the movement.

Therefore, it is not known if the time spent producing the

gestures were abnormally long in the right brain-damaged

group, which would have suggested a continuous

perseveration.

As was stated above, neither brain-damaged group

demonstrated significant differences in external

configuration. This is not consistent with what has

previously been reported. Specifically, Haaland and

Flaherty (1984) reported that when the task involved