Академический Документы
Профессиональный Документы
Культура Документы
www.elsevier.com/locate/cogbrainres
Research Report
Moss Rehabilitation Research Institute, Korman 213, 1200 W. Tabor Road, Philadelphia, PA 19141, USA
b
Thomas Jefferson University, Philadelphia, PA 19107, USA
Accepted 27 May 2005
Available online 5 July 2005
Abstract
A considerable recent literature argues that the same representations, encoded by inferior prefrontal and parietal cells known as mirror
neurons, may be activated in both production and recognition of object-related actions. Here, we test several predictions derived from the
contemporary literature on the parity between production and recognition and the putative emergence of the mirror neuron system from a system
coding hand object interactions. Forty-four patients with left-hemisphere stroke, 21 of whom exhibited ideomotor apraxia, performed a number
of pantomime imitation and recognition tasks, and performance was scored with respect to hand posture, arm posture, amplitude, and timing.
Consistent with predictions, there were strong relationships between object-related pantomime imitation and object-related pantomime
recognition, and between imitation and recognition of the hand posture component of object-related actions. Skilled object-related gesture
representations are likely to be closely tied to evolutionarily more primitive systems controlling object grasping, to emerge from a mapping
between object and action information coded by ventral and dorsal streams, and to be lateralized to the left hemisphere in humans.
D 2005 Elsevier B.V. All rights reserved.
Theme: Neural basis of behavior
Topic: Cognition
Keywords: Apraxia; Action; Gesture; Imitation; Recognition
1. Introduction
According to the direct matching hypothesis, actions
performed by others are recognized by activating the same
spatiomotor representations used for performing the action
oneself. Numerous recent investigations in infant development (e.g., [37]) and adult cognitive psychology (e.g.,
[4,47]) suggest that there is a common coding between
perception and action. A possible physiological foundation
for at least some aspects of this common coding is provided
by the recent discovery of so-called mirror neurons in the
* Corresponding author. Moss Rehabilitation Research Institute, Korman
213, 1200 W. Tabor Road, Philadelphia, PA 19141, USA. Fax: +1 215 456
5926.
E-mail address: Lbuxbaum@einstein.edu (L.J. Buxbaum).
0926-6410/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.cogbrainres.2005.05.014
227
228
2. Subjects
Forty-four left-hemisphere stroke patients participated in
the study. All patients had suffered a single left-hemisphere
cerebral vascular accident; one subject had an additional
small asymptomatic right occipital infarct pre-dating the
left-hemisphere stroke. All subjects gave informed consent
to participate in accordance with guidelines of Albert
Einstein Healthcare Network and were paid for their
participation. Subjects were referred to the study from a
large database of potential research subjects in the Philadelphia area maintained by Moss Rehabilitation Research
Institute. Subjects were excluded if database records
indicated language comprehension deficits of sufficient
severity to preclude comprehension of task instructions.
Subjects over the age of 80 and/or with histories of comorbid neurologic disorders, alcohol or drug abuse, or
psychosis were also excluded. All subjects gave informed
consent to participate in accordance with the guidelines of
the IRB of Albert Einstein Healthcare Network, and were
paid for their participation.
We pursued two complimentary strategies in analyzing
the data. The first strategy was to classify participants as
apraxic or not, and assess whether performance of the two
groups differed on measures of interest. The second strategy
was to treat data as continuous variables to assess the
strength of relationships of scores on the measures of
interest.
Subjects were characterized as exhibiting ideomotor
apraxia (hereafter, IM) or not (hereafter, LCVA) on the
3. Language comprehension
Participants performed the Comprehension subtest of the
Western Aphasia Battery (Kertesz Ref). Maximum possible
score was 200 points. Apraxics exhibited somewhat greater
deficits in comprehension (mean = 166, range = 117 196)
than did LCVA (mean = 187, range = 111 200), t(42) = 3.1,
P < 0.01.
4. Lesion analysis
Clinical T-1 or T-2-weighted MRI scans were available
for 36 of the 44 subjects (18 LCVA and 18 IM). Lesions
were segmented and interpreted by an experienced neurologist. Subtractions of the lesioned regions of the IM versus
LCVA groups were performed by one of the authors using
the MRIcro image analysis program developed by Dr. Chris
Rorden (see http://www.psychology.nottingham.ac.uk/staff/
cr1/mricro.html). Fig. 1 shows the result of this subtraction
229
Table 1
Subject demographics and scores on gesture production and imitation
Subject Group
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
A11
A12
A13
A14
A15
A16
A17
A18
A19
A20
A21
L1
L2
L3
L4
L5
L6
L7
L8
L9
L10
L11
L12
L13
L14
L15
L16
L17
L18
L19
L20
L21
L22
L23
Gesture to sight of object Meaningless imitation Gesture composite Age Education Gender Handedness Lesion volume (cm3)
IM
64
IM
75
IM
60
IM
45
IM
75
IM
55
IM
38
IM
73
IM
63
IM
65
IM
63
IM
75
IM
70
IM
50
IM
65
IM
60
IM
55
IM
75
IM
50
IM
65
IM
58
LCVA 90
LCVA 85
LCVA 88
LCVA 88
LCVA 90
LCVA 90
LCVA 97
LCVA 85
LCVA 93
LCVA 88
LCVA 88
LCVA 90
LCVA 90
LCVA 88
LCVA 95
LCVA 100
LCVA 88
LCVA 93
LCVA 83
LCVA 93
LCVA 78
LCVA 85
LCVA 90
60
58
40
73
75
25
50
58
58
75
63
58
68
78
58
50
90
58
53
50
65
90
93
90
93
93
93
85
90
93
93
95
93
95
100
95
93
93
93
90
88
95
85
98
62
66
50
59
75
40
44
65
60
70
63
66
69
64
61
55
73
66
51
58
61
90
89
89
90
91
91
91
88
93
90
91
91
93
94
95
96
90
93
86
90
86
85
94
55
49
79
79
50
59
63
67
56
79
64
50
49
42
60
78
42
67
39
58
41
35
56
51
64
77
42
55
51
58
58
50
65
77
50
51
80
55
56
54
42
54
40
64
18
16
12
11
12
16
12
12
14
16
12
10
8
16
12
11
16
14
10
10
10
12
20
14
12
12
15
12
12
12
3
12
19
16
18
8
12
16
12
12
8
18
16
16
F
F
F
F
M
M
F
F
M
F
M
F
M
F
M
M
F
M
M
F
F
F
M
F
M
M
F
M
F
M
M
M
M
F
F
M
F
F
M
F
M
M
F
M
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
L
R
R
R
R
R
L
R
R
208.4
110.5
58.2
8.5
64.3
68.7
253.5
7.2
44.9
NA
NA
151.9
49.0
143.6
161.5
51.1
131.3
119.7
180.8
45.3
46.1
41.8
96.8
56.7
77.4
NA
16.8
69.3
41.4
58.6
0.5
NA
61.6
13.6
31.2
28.0
0.4
18.4
22.4
NA
178.0
95.1
142.2
25.6
5. Experimental tasks
5.1.1. Methods
Participants watched videotapes of an examiner performing 10 transitive and 5 intransitive pantomimes with the
right hand, and were asked to imitate the gesture as precisely
as possible with the unimpaired left hand. Transitive
gestures were hammering, cutting with scissors, sawing,
using a screwdriver, writing with a pencil, using a comb,
winding a watch, brushing teeth, flipping a coin, and eating
with a fork. Intransitive gestures were saluting, waving
goodbye, hitch-hiking, signaling stop, and beckoning
230
Fig. 1. Subtractions of lesioned regions of IM versus LCVA groups. These indicate the difference in the proportion of patients in the two groups having
involvement in a given region. Colors further to right on the color bar indicate relatively more lesion in IM group in increments of 20%. Left and right are
reversed. There were no regions uniquely damaged in the IM group; however, there were several regions in which lesions were more likely. These include
Brodmann areas 6 and 44 (dorsolateral frontal), 22 and 37 (temporal), and 39 and 40 (inferior parietal).
5.1.2. Results
Scores are reported here in terms of percent correct: IM
transitive mean = 63%, intransitive mean = 89%, mean
transitive intransitive difference = 25.6%; LCVA transitive mean = 93%, intransitive mean = 99%, mean
difference = 4.7%; CTL transitive mean = 93%, intransitive
mean = 100%, mean difference = 5.8%. In this and all
subsequent analyses, the difference between transitive and
intransitive performance was calculated for each subject,
and between-group comparisons of the difference scores
were performed with Kruskal Wallis nonparametric oneway ANOVAs. Post hoc testing was performed with Mann
Whitney tests. There was an effect of group, H = 28.36, P <
0.0001, indicating that the transitive intransitive difference
scores were greater for some group(s) than others. Post hoc
testing with the critical value of P set at .016 (i.e., 0.05/3)
confirmed that the difference scores were significantly
greater for the IM patients than for the other two groups
(Ps < 0.0001), who did not differ from one another (P >
0.36). These data replicate previous reports that objectrelated gestures are more difficult for IM patients than are
symbolic gestures.
In the next analysis, we assessed whether IM patients
were equally impaired in all components of transitive
gesture. Fig. 2 shows the data entered into these analyses.
3
To assess scoring reliability, gestures for 6 of the participants were
scored by 2 independent coders. Percent agreement between the coders
ranged from 78% to 100% across the 6 subjects (mean 88% agreement;
Cohens kappa = 0.60).
intransitive performance. There were significant betweengroup differences in hand posture, H = 35.2, P < 0.0001;
arm posture, H = 10.5, P = 0.003; and amplitude, H = 12.3,
P = 0.002. Post hoc testing of the 3 significant ANOVAs
with Mann Whitney tests using a Bonferroni-corrected P
value of .0055 (i.e., 0.05/9) indicated that there were
significant differences between the IM patients and the
other two groups for hand posture and amplitude (P <
0.0001), and a difference between IM and LCVA for arm
posture ( P = 0.004). None of the differences between LCVA
and CTL were significant (all Ps > 0.27). Thus, the
disproportionate disparity between transitive and intransitive gesture in the IM group was reliably observed in several
gesture components.
In a final analysis, we addressed the concern that the
disproportionate impairment in transitive as compared to
intransitive hand postures in the IM group may be related
to the greater complexity of the former. We examined the
data from the IM and LCVA groups for a subset of 5
transitive gestures having a simple, stable hand posture
denoting grasp of a tool (hammering, sawing, combing
hair, brushing teeth, and eating with a fork) and all 5 of the
intransitive gestures examined previously (waving goodbye, beckoning come here, hitch-hiking, signaling stop,
and saluting). Within-group comparisons were performed
with Wilcoxon Signed Ranks Tests (with a Bonferronicorrected P value of .008, i.e., .05/6, required for significance). For transitive movements, IM patients hand
posture (mean 50.4% correct) tended to be more deficient
than arm posture (66.7%; P = 0.02), amplitude (69%; P =
0.01), or timing (82%; P = 0.009). For intransitive
imitation, IM patients hand posture (mean 91% correct)
tended to be better than arm posture (84%; P = 0.01), and
was equal to amplitude (91%) and timing (89.5%). In
between-group comparisons, Mann Whitney U tests (with
a Bonferroni-corrected P value of 0.016 required for
significance) confirmed that the disparity between transitive
and intransitive hand postures was more pronounced for the
IM patients than the other two groups (Ps < 0.008), who
did not differ from one another (P = 0.5).
5.1.3. Discussion
In this study, we replicated previous findings indicating
that transitive gesture imitation is more impaired in patients
with IM than is intransitive imitation. Moreover, consistent
with the possibility that the system that is damaged in IM
patients is specialized for hand object interactions, the hand
posture component of gesture imitation for transitive
(object-related) gestures proved to be the most impaired
aspect of IM patients performance.
One possible objection to the proposed interpretation is
that transitive hand postures may simply be more difficult
than other components of transitive gesture, and more
difficult than intransitive hand postures, and thus more
sensitive to any type of impairment. There are two lines of
evidence against this interpretation. First, when we com-
231
232
233
Imit.
Imit.
Imit.
Imit.
Imit.
Imit.
T P < 0.0028.
TT P < 0.001.
TTT P < 0.0001.
Recog. trans
hand posture
Recog. trans
arm posture
Recog. trans
amp/time
0.80TTT
0.77TTT
0.58TT
0.38
0.57TT
0.49T
0.48T
0.52TT
0.25
0.40
0.40
0.21
0.37
0.33
0.29
0.35
0.43
0.30
234
5.3.1. Discussion
The data from three analyses of the relationship between
gesture production and recognition tell a consistent story.
The first analysis showed that transitive gesture imitation is
a strong and unique predictor of transitive gesture recognition. The second analysis showed that individual components of transitive gesture recognition (hand posture, arm
posture, and amplitude/timing) are more strongly related to
these same components instantiated in transitive as compared to intransitive gesture imitation. The third analysis
showed that patients who fare poorly in transitive gesture
recognition are disproportionately impaired in transitive (as
compared to intransitive) imitation, whereas patients who
are better at transitive gesture recognition show less
disparity between the two types of imitation. Together,
these data argue that the same representations subserve the
recognition and imitation of transitive gesture. Furthermore,
they suggest that these representations may be componential, with hand posture representations for transitive actions
particularly vulnerable to disruption in left-hemisphere
stroke (but not in CBD, as noted earlier). In the General
discussion section, we will present a model that addresses
what may be special about these representations for
transitive gesture and how they may differ from intransitive
gesture representations.
5.4.1. Methods
Patients lesions were segmented by a neurologist (H. B.
Coslett) and drawn into MRIcro software by an experienced
physician (R. Menon). Brodmann areas were identified by
the first author and Dr. Coslett, who were blinded to
subjects identities, using templates from Damasio and
Damasio [11] and Mai and Assheuer [36].
Patients were ranked in terms of their performance on
the Spatial gesture recognition task, the hand posture score
from the gesture recognition task, the total score from the
transitive gesture imitation task, and the hand posture score
from the transitive gesture imitation task (4 separate
rankings). For each measure, we divided the patient group
into thirds (insofar as permitted due to ties) and discarded
the patients whose performance was in the central third of
the distribution. The lesion data from the high and low
performing groups on the recognition task are shown in
Fig. 5.
We used MRIcro software to identify whether a lesion
was present or absent in 5 Brodmann areas of interest: areas
44 and 45 (inferior prefrontal) and areas 39, 40, and 7
(posterior parietal lobe and intraparietal sulcus) in the high
versus low performing group. Lesions in BA 39 (angular
gyrus) and the inferior portion of area 7, including the
superior bank of the intraparietal sulcus, were more
frequently associated with low Spatial gesture recognition
scores than high Spatial recognition scores v 2 > 4.6, P <
0.02 for both regions), with low hand posture recognition
scores than high hand posture recognition scores (v 2 > 4.3,
P < 0.02 for both), and with low total gesture imitation
scores than high total gesture imitation scores (v 2 > 5.0, P <
0.02 for both). Lesions in the inferior portion of area 7,
including the intraparietal sulcus, were also significantly
more strongly associated with low hand posture imitation
scores than with high hand posture imitation scores (v 2 =
5.6, P = 0.01). None of the comparisons involving areas 44
or 45 even approached significance (all Ps > 0.2).
5.4. Neuroanatomy
A final question regards the neuroanatomic basis of
transitive gesture representations, and particularly the
component of those representations that is specialized for
hand object interactions. As noted, mirror neurons active
5.4.2. Discussion
Lesion analyses suggest that in stroke patients, the
lesion(s) significantly associated with deficits in the
recognition of transitive gesture, and the hand posture
component of transitive gesture, are located in the inferior
parietal lobe and intraparietal sulcus. For prefrontal cortex,
in contrast, the association with deficits in gesture recognition or imitation did not even approach significance. This
is at least partially consistent with earlier findings of Varney
and Damasio [60] indicating that patients with deficits in
matching pantomimes to associated objects were likely to
have lesions in area 40 (supramarginal gyrus), areas 22 and
235
Fig. 5. Subtracted lesioned regions in high versus low performing groups on the Spatial Recognition task (top left and right) and Hand Posture score of the
Spatial Recognition task (bottom left and right). See Fig. 1 for additional explanation.
6. General discussion
The hypothesis that IM reflects a deficit in the
representations underlying skilled object-related gestures,
with particular degradation of the hand posture component
of skilled object-related gestures, enabled us to generate a
number of predictions that were tested in the present study.
We demonstrated that patients with IM were disproportionately impaired in the imitation of transitive as compared to
intransitive gestures, and were particularly impaired in the
236
237
Acknowledgments
Supported by NIH RO1-NS36387 and NIDRR H133G0
30169 to the first author, and by the Pennsylvania Department of Health. The Department of Health specifically
disclaims responsibility for any analyses, interpretations, or
conclusions. We are grateful to Branch Coslett for performing lesion analyses and Michael Arbib for his thoughtful
comments on an earlier draft of the manuscript.
1. Hand Posture
Score as 0 if hand posture/grasp is unrecognizable,
flagrantly incorrect, or only transiently correct (small
fragment of total gesture with correct posture or
grasp). Score 0 for Fbody part as object_ (BPO)
errors.
Score as 1 if posture is correct or subtly incorrect
(e.g., hand aperture slightly too big or small; wrist
angle slightly incorrect).
2. Arm Posture/Trajectory
Score as 0 if arm posture and/or trajectory (e.g.,
joint angles, plane of movement relative to body/
environment (e.g., side to side instead of back and
forth), shape of movement (e.g., circular instead of
linear) are flagrantly incorrect or only transiently
correct (small fragment of total gesture with correct
posture).
Score as 1 if both arm posture and trajectory are
correct; or if arm posture and/or trajectory are subtly
incorrect (e.g., elbow slightly too bent, trajectory at
slight angle relative to what is appropriate, shape of
movement slightly distorted).
3. Amplitude
Score as 0 if size of movement is clearly too large or
too small (e.g., sawing with small scratching
movement), or if size is only transiently correct (small
fragment of total gesture with correct amplitude).
Score as 1 if size is correct or subtly too large or too
small (e.g., slight Fovershoot_ or Fundershoot_ in
movement amplitude).
238
4. Timing/Frequency
Score as 0 if speed of movement is flagrantly too
fast or slow; and/or if number of cycles of movement
is flagrantly too few or many (e.g., Fflipping_ coin 4
times in succession; Fscissoring_ only once).
Score as 1 if speed of movement is subtly too fast or
slow; and/or if frequency is subtly inappropriate (e.g.,
flipping coin twice; scissoring only twice).
References
[1] C. Barbieri, E. DeRenzi, The executive and ideational components of
apraxia, Cortex 24 (4) (1988) 535 543.
[2] B. Bell, Pantomime recognition impairment in aphasia: an analysis of
error types, Brain Lang. 47 (1994) 269 278.
[3] G. Buccino, F. Binkofski, G.R. Fink, L. Fadiga, L. Fogassi, V. Gallese,
R.J. Seitz, K. Zilles, G. Rizzolatti, H.J. Freund, Action observation
activates premotor and parietal areas in a somatotopic manner: an
fMRI study, Eur. J. Neurosci. 13 (2001) 400 404.
[4] G. Butterworth, On reconceptualizing sensori-motor coordination in
dynamic system terms, in: H.B.B.I. Bertenthal (Ed.), Sensory Motor
Organization and Development in Infancy and Early Childhood,
Kluwer Academic Press, The Netherlands, 1990, pp. 57 73.
[5] L.J. Buxbaum, Ideomotor apraxia: a call to action, Neurocase 7 (2001)
445 458.
[6] L.J. Buxbaum, T. Giovannetti, D. Libon, The role of the dynamic body
schema in praxis: evidence from primary progressive apraxia, Brain
Cogn. 44 (2000) 166 191.
[7] L.J. Buxbaum, A. Sirigu, M.F. Schwartz, R.L. Klatzky, Cognitive
representations of hand posture in ideomotor apraxia, Neuropsychologia 41 (2003) 1091 1113.
[8] L.J. Buxbaum, S.H. Johnson-Frey, M. Bartlett-Williams, Deficient
internal models for planning hand object interactions in apraxia,
Neuropsychologia 43 (2005) 917 929.
[9] L.J. Buxbaum, K. Kyle, M. Grossman, H.B. Coslett, Left inferior
parietal representations for skilled hand object interactions: evidence
from stroke and corticobasal degeneration, Cortex (in press).
[10] C. Cavada, The visual parietal areas in the macaque monkey: current
structural knowledge and ignorance, NeuroImage 14 (2001) S21 S26.
[11] H. Damasio, A.R. Damasio, Lesion Analysis in Neuropsychology,
Oxford Univ. Press, New York, NY, 1989.
[12] E. DeRenzi, Methods of limb apraxia examination and their bearing
on the interpretation of the disorder, in: E.A. Roy (Ed.), Advances in
Psychology: Neuropsychological Studies of Apraxia and Related
Disorders, vol. 23, North Holland, Amsterdam, 1985.
[13] C. Distler, D. Boussaoud, R. Desimone, L.G. Ungerleider, Corical
connections of inferior temporal area TEO in macaque monkeys,
J. Comp. Neurol. 334 (1993) 125 150.
[14] J.R. Duffy, R.J. Duffy, Three studies of deficits in pantomimic
expression and pantomimic recognition in aphasia, J. Speech Hear.
Res. (1981) 70 84.
[15] C. Dumont, B. Ska, Pantomime recognition impairment in Alzheimers
disease, Brain Cogn. 43 (2000) 177 181.
[16] D. Eidelburg, A.M. Galaburda, Inferior parietal lobule: divergent
architectonic asymmetries in the human brain, Arch. Neurol. 41
(1984) 843 852.
[17] A. Fagg, M.A. Arbib, Modeling parietal premotor interactions in
primate control of grasping, Neural Netw. 11 (1998) 1277 1303.
[18] J.M. Ferro, I.P. Martins, G. Mariano, A.C. Caldas, CT scan correlates
of gesture recognition, J. Neurol., Neurosurg. Psychiatry 46 (1983)
943 952.
[19] G. Gainotti, M. Lemmo, Comprehension of symbolic gestures in
aphasia, Brain Lang. 3 (1976).
239