Deconstructing apraxia: understanding disorders of intentional

movement after stroke

Lisa Koskia,b, Marco Iacobonia,c and John C. Mazziottaa,b,d,e

Impairments in praxic functioning are common after stroke, most Introduction

frequently when the left hemisphere is affected. Several recent Apraxia is dened as a decit in the ability to understand
studies of apraxia after stroke have made advances in an action or to perform an action in response to verbal
understanding the right hemisphere contribution to praxis, command or in imitation (e.g. wave goodbye, pantomime
particularly for the performance of novel actions. Moreover, use of a hammer), in the absence of basic sensory or
quantitative lesion analysis in stroke patients indicates the motor impairments. This decit may be conceptualized
importance of cortical regions such as the intraparietal sulcus as a decit in the mental representation of specic
and the middle frontal gyrus for subserving praxic function. aspects of an action. We will use the term apraxia here,
Complex neuropsychological models have been developed to although most patients show only partial effects and
account for the many dissociations observed in the types of might therefore be considered dyspraxic, rather than
errors observed in stroke patients. Relatively lacking, however, apraxic. Although apraxia is most commonly associated
are models that attempt to relate the neurological data to what is with strokes affecting the left parietal lobe, it may also
known about praxis from functional neuroimaging in normal occur in lesions to the right parietal lobe, the temporal or
subjects and from physiological studies in the monkey. frontal lobes, and even subcortical regions including
Moreover, a coherent interpretation of the results of apraxia white matter and the basal ganglia [1 .]. Apraxia and
studies remains hampered by the lack of a standard testing aphasia often occur together, but apraxia can occur in the
instrument to assess the nature and severity of apraxic absence of aphasia and is, therefore, considered a distinct
impairments in the groups tested. Curr Opin Neurol 15:7177. # 2002 disorder [2].
Lippincott Williams & Wilkins.
Here, we present an overview of current models of
apraxia followed by a review of the studies of apraxia
published in the neurological literature over the past
a
Ahmanson-Lovelace Brain Mapping Center, Neuropsychiatric Institute, Departments year. We also provide an analysis of the neuroimaging
of bNeurology, cPsychiatry and Biobehavioral Sciences, dPharmacology and literature from the past year that is relevant to the
e
Radiological Sciences, UCLA School of Medicine, Los Angeles, USA
understanding of normal and impaired praxis.
Correspondence to Lisa Koski, Ahmanson-Lovelace Brain Mapping Center, 660
Charles E. Young Drive South, Los Angeles, CA 90095-7085, USA
Tel: +1 310 794-4964; fax: +1 310 794-7406; e-mail: [email protected] Models of apraxia
The earliest theories of apraxia formulated by Liepmann
Current Opinion in Neurology 2002, 15:7177
postulated that the left parietal cortex is important for
# 2002 Lippincott Williams & Wilkins
the formation of motor programs that specify the spatial
1350-7540 and temporal sequence of movements composing an
action [3,4]. Ideational apraxia would arise when the
motor programming area is destroyed by damage to the
supramarginal gyrus, impairing the conceptual represen-
tation of an action and leading to decits in using tools or
performing an action to verbal command while imitation
is spared. Performance of meaningful action sequences
(e.g. lighting a candle) was also used as a test for
ideational apraxia. Ideomotor apraxia would arise when
the motor programming area is disconnected from the
premotor and motor regions, so that the patient can
conceptualize but not actually execute the action,
demonstrating spared recognition of tools but decient
ability to use them appropriately or to imitate actions.

Since that time, a great number of terms have evolved

for classifying different subtypes of apraxia based on
subtle dissociations in praxic abilities. The result is
confusion in the literature in which the same term may
71
72 Cerebrovascular disease

Table 1. Glossary of apraxia terms sequencing, omission and perseverative errors viewed as
Type of apraxia Definition characteristic of ideational apraxia, whereas single-object
tests (e.g. using a key) were intended to elicit the
Buccofacial apraxia Impairment in performing mouth or face actions
on verbal command or imitation (see also spatiotemporal errors in tool use that characterized
orofacial apraxia). ideomotor apraxia. Correlations and factor analyses of
Conceptual apraxia Form of apraxia in which the concept of the these two tasks within the same group of patients,
action is lost; characterized by impaired ability to
use tools and to understand meaningful however, indicate that both appear to involve similar
gestures. underlying processes [6 .]. Moreover, multiple-object
Constructional apraxia Inability to assemble component parts into a actions may be performed more easily than single-object
coherent whole
Ideational apraxia Impairment in the sequential use of multiple actions by some apraxic patients, and even by normal
objects. Traditionally, also used to refer to controls, because the association between related objects
impairment in the concept of an action. The term may cue correct performance.
conceptual apraxia (see above) was coined to
distinguish between these two dissociable
impairments. A wealth of case reports have demonstrated dissociations
Ideomotor apraxia Impairment in the performance of skilled in the ability of individual patients to recognize versus
movements on verbal command or in imitation;
most commonly characterized by spatial or use tools, imitate meaningful versus meaningless actions,
temporal errors in movement execution. perform transitive versus intransitive actions, and per-
Limb apraxia Usually used to refer to ideomotor apraxia of the form limb versus orofacial actions [7,8 .]. These observa-
limbs; frequently includes impaired performance
of actions that also depend on the hands and tions are interpreted as reecting modularity in the
fingers. praxis system and a number of fairly complex models of
Limb-kinetic apraxia Slowness and stiffness of movements with apraxia have now been developed (Fig. 1). Separate
a loss of fine and precise movements.
Optical apraxia Impairment in performing saccadic eye semantic and non-semantic pathways allow for dissocia-
movements on command tions in the ability to represent meaningful versus
Orofacial apraxia Impairment in performing mouth or face actions meaningless actions; separate input and output lexicons
on verbal command or imitation (see also
buccofacial apraxia). account for differences in the ability to conceptualize
Speech apraxia Selective impairment in ability to produce actions and the ability to perform them; and separate
speech sounds input pathways for verbal and visual stimuli explain the
Tactile apraxia Impairment of hand movements for the use of
and interaction with an object, in the presence of dissociation between the ability to perform an action on
preserved intransitive movements. command versus in imitation.
Unimodal apraxia Any form of apraxia that is specific to actions
demonstrated in a single modality, e.g. visual,
but not auditory. Despite the ability of the model depicted in Fig. 1 to
account for the dissociations in apraxic impairments
observed in patients, some features deserve further
scrutiny. First, the model was developed from the study
be used differently by different investigators. Table 1 of ideomotor apraxia of the limb and, thus, emphasizes
provides the most commonly accepted denitions for a praxis of meaningful actions. The non-semantic pathway
variety of different disorders labelled as apraxia. was included in the traditional models because mean-
ingful actions can be imitated without awareness of their
The most commonly used term now is ideomotor meaning. Indeed, the terms `semantic' and `non-
apraxia. In its strictest sense, ideomotor apraxia refers semantic' were borrowed from the aphasia literature.
to the impaired reproduction of meaningful or learned However, it may be more accurate to refer to meaningful
actions, although the ability to perform or imitate actions as well-practised or learned actions that may be
meaningless actions is also decient in some cases. already represented in their entirety in an individual's
The use of the term ideational apraxia is less common motor repertoire, and to meaningless actions as actions
now than when it was rst conceived due to difculties involving novel motor sequences that must be analyzed
in establishing it as a valid and coherent entity distinct and constructed from existing movements. In this
from the symptoms ascribed to ideomotor apraxia or context, it is clear that the non-semantic pathway, often
decits in executive functioning. Researchers soon saw a termed the `direct' pathway, is notably underspecied in
need to discriminate between sequencing errors, re- these models and may be far from direct. In fact,
ferred to in the traditional way as ideational apraxia, and researchers studying apraxia for meaningless actions are
impairments in action recognition, termed `conceptual discovering that even novel or meaningless actions with
apraxia' [5]. Recent work comparing impairments in no pre-learned representations may be represented at a
single-object and multiple-object use provides another conceptual level under certain conditions [9,10 . .]. It is
illustration of the difculties with the traditional concept not yet known whether the conceptual representation of
of ideational apraxia. Tests of multiple-object use (e.g. meaningless actions would use some of the same
lighting a candle) were originally designed to elicit pathways as the representation of meaningful actions
 Apraxia after stroke Koski et al. 73

Figure 1. Standard model of limb praxis Figure 2. Revised model of limb praxis, inwhich dynamic body-
schema representations interact with stored representations of
learned actions

Verbal Visual object Visual gesture

stimuli stimuli stimuli
Verbal Visual object Visual gesture
stimuli stimuli stimuli

Phonological Visual
analysis analysis Spatiomotor transformations
from retinotopic to
intrinsic body centered
coding
Phonological Structural
description Action input Knowledge of tools
input lexicon system lexicon characteristic movements
Dynamic model of body
coded in intrinsic
spatial coordinates
Stored portion of (body schema)
Action gesture representation
semantic
system
Response

Derived from Buxbaum et al. 2000 [11 . .].

Action output
lexicon

suggests that the brain may represent observed actions

Gestural
Visuomotor by mapping them directly onto the same substrates that
conversion
buffer mechanism are used to execute an action. Recent work using
functional magnetic resonance imaging and magneto-
encephalography has demonstrated the existence of a
Response similar mirror system in the human brain [13,14].

Derived from Rothi et al. 1991 [7] and Cubelli et al. 2000 [8 .]. Lesion correlates of apraxia:
interhemispheric
Roughly 30% of patients in the acute phase of stroke
and whether it would require an expansion of the `direct' show evidence of apraxia [3,15], but the incidence is
pathway. A recent model of the praxis system proposed higher after damage to the left hemisphere (50%) than to
by Buxbaum and colleagues (see Fig. 2) provides a very the right hemisphere (510%) [16]. Nevertheless, con-
different way of describing the relationship between siderable variability in this estimate is found across
meaningful and meaningless actions. This improved studies due to the lack of standardized assessment tools
model proposes an interplay between a dynamic body- and wide variations in criteria for diagnosing the
centred representation of actions and stored representa- disorder.
tions of learned actions [11 . .]. From a physiological
perspective, these changes improve the plausibility of Several recent studies have focused on whether
the model, although there is still no attempt to explain different patterns of apraxic impairment are associated
the possible relationships between the functions repre- with left hemisphere damage or right hemisphere
sented and the neuropsychological mechanisms that damage. It appears that comparable levels of impair-
underlie them. ment may be seen across both groups for certain types
of apraxic impairment. In a study of facial apraxia after
Second, the separation of the input and output lexicons stroke, patients with left hemisphere damage made
may prove untenable in light of the recent discovery of a more errors than patients with right hemisphere
so-called mirror system in the ventral premotor cortex damage when imitating lower face actions, whereas
and the posterior premotor cortex of the monkey brain. both groups performed similarly for upper face actions
Neurons in area F5 of the premotor cortex tend to alter [17 .]. In a study of limb apraxia, patients with left
their pattern of ring during the performance of a hemisphere strokes were more likely than patients
specic action, such as precision grip, but also show with right hemisphere strokes to be impaired at
additional response properties. A subset of neurons in pantomiming the use of a tool in response to verbal
this region, the mirror neurons, also respond to the mere command, whereas an equal proportion of left hemi-
observation of the represented action performed by sphere and right hemisphere patients were impaired
someone else [12]. The existence of mirror neurons when imitating a pantomime demonstrated by the
74 Cerebrovascular disease

experimenter [18 .]. Although half of the patients with precentral lesions were not impaired on the recognition
left hemisphere stroke also had aphasia, the correlation or performance of unimanual pantomimes, but those
with apraxia was similar for the pantomime and with damage to medial premotor regions were severely
imitation conditions. impaired at pantomiming bimanual actions that required
different movements from each hand. The importance of
These ndings suggest an important role for the right medial premotor regions for bimanual actions is con-
hemisphere in praxis, perhaps more particularly for novel sistent with neuroimaging evidence in normal subjects of
actions. The results of studies in which patients imitated disproportionately increased activity in supplementary
meaningless actions may provide a direct test of this motor area during bimanual compared with unimanual
hypothesis. A recent investigation of patients with actions [25].
relatively circumscribed lesions to the parietal lobe found
that only those with left hemisphere damage are impaired A recent study used quantitative structural image
at performing meaningless actions on verbal command or analyses to determine the location of greatest lesion
imitation [19]. Unfortunately, the interpretation of this overlap in patients with anterior or posterior strokes. The
study is problematic since the left hemisphere group was intraparietal sulcus and the middle frontal gyrus were
composed primarily of stroke patients whereas the right identied as the regions most important for the imitation
hemisphere group had tumours, which may have allowed of actions [26 .]. The use of structural scans to report the
for long-term reorganization of function. Recent studies extent and overlap of lesions in groups of patients is a
by Goldenberg and colleagues, however, provide persua- relatively new feature of this study that represents an
sive evidence of the contribution of the right hemisphere improvement in the analysis of lesiondecit relation-
to the recognition and imitation of meaningless actions. ships. The validity of this approach is demonstrated by
Patients with left hemisphere strokes were impaired at the similarity between the brain regions identied in this
imitating all meaningless actions and at perceptual study and those identied in functional neuroimaging
matching of meaningless hand-to-head actions [9,20]. In studies of imitation in normal subjects [12].
contrast, patients with right hemisphere lesions were
impaired at imitating meaningless postures of the ngers Functional imaging of praxis
of the hand and at perceptual matching of all meaningless Functional neuroimaging studies conrm the importance
actions. Consistent with these ndings are the results of a of the parietal and frontal cortex in the representation of
recent study in a patient with callosal resection, in whom actions under a variety of conditions including action
the left hemisphere initially had difculty supporting observation, action recognition, observation of graspable
imitation of nger postures on its own and neither objects, imagined action, and imitation [27]. A current goal
hemisphere could support perceptual matching of nger of such studies is to determine whether the contributions
postures when working on its own. [21]. From this work, of the parietal and frontal cortex differ, and whether there
it may be concluded that the left hemisphere is important are specic regions within the parietal or frontal cortex that
for representing actions in terms of knowledge about the contribute preferentially to particular aspects of praxis.
structure of the human body, whereas the right hemi-
sphere participates in the visuospatial analysis of ges- To date, the results of neuroimaging studies support the
tures. This suggests that apraxia after right hemisphere notion that both hemispheres contribute to praxis,
strokes results from disruption of a pathway for trans- although the left hemisphere may be slightly more
lating visual input to motor output, which preferen- involved for some tasks. The left hemisphere, particu-
tially impairs the representation of novel actions [10 . .]. larly the inferior parietal lobule, is more active than the
right during perceptual discrimination of meaningless
Lesion correlates of apraxia: gestures, but when the gestures consist of nger
intrahemispheric postures, the intraparietal sulcus of the right hemisphere
Although some studies have not found an association shows greater activity [28 .]. Two recent studies found
between the locus of the lesion within a hemisphere and greater activity in the left inferior parietal lobe compared
the severity of apraxia [22,23] others are now focusing on with the right parietal lobe for recognition of transitive
whether lesions to particular subregions within each actions [29] and tools [30], but these studies used
hemisphere may be associated with particular subtypes naming tasks and were therefore biased to nding left
or patterns of apraxic performance. Halsband and hemisphere effects. Neuroimaging studies of action
colleagues [24 .] demonstrated that patients with parietal imagery, imitation, or observation found increased
lesions show the most severe impairments in the activity in the left [31], right [13] and bilateral [32 .]
recognition and imitation of pantomimed actions and posterior parietal cortex, respectively. Although in-
the decit is particularly severe for left parietal lesions creased signal in premotor cortex is more commonly
and actions directed toward their own bodies (e.g. lateralized to the left hemisphere, it has been observed
brushing hair). In contrast, patients with premotor or bilaterally in some studies [32 .].
 Apraxia after stroke Koski et al. 75

Besides providing converging evidence for the neural work after a stroke [40]. Therefore, a behavioural
mechanisms underlying apraxia, neuroimaging studies training programme focused on the imitation of actions
may also provide new insights into how actions are has been developed that appears to produce improve-
represented in the brain and how such representations ments specic to tests of apraxia in left hemisphere
are modulated by specic task demands. A recent stroke patients [41].
functional magnetic resonance imaging study of action
observation suggested that actions may be somatotopi- Another therapeutic approach is to teach strategies to
cally represented within the parietal and premotor compensate for continued apraxic decits. Using such an
cortex, with mouth actions being represented ventrally, approach improvements were reported in activities of
foot actions more dorsally, and armhand actions in daily living, but not on tests of apraxia [42,43].
between [32 .]. Peigneux and colleagues have recently Unfortunately, these studies did not include a control
embarked on an ambitious project using positron group for pre- and post-treatment assessment so it is
emission tomography to map each of the components impossible to determine whether the improvements
specied by a cognitive model of praxis processing observed were the result of the intervention or part of
similar to that shown in Fig. 1. Published results to date the natural course of recovery.
support the need for elaboration of the `non-semantic'
pathway to involve dorsal parietal regions important for Conclusion
visuospatial processing [33,34]. There appears to be a consensus within the literature
that the left parietal cortex subserves a particularly
Little is known about the neural functioning of patients important component of the praxis system, especially
with apraxia, although two case studies showed hypo- concerned with the knowledge or representation of
metabolism of the left parietal lobe [35,36]. A recent overlearned actions. It is recognized, however, that
electrophysiological study measured patterns of slow damage to cortical and/or subcortical regions outside
cortical potentials and of event-related desynchonization the left parietal cortex, including the right hemisphere,
of activity in the alpha and beta bands in groups of stroke have also been associated with apraxia and it is assumed
patients with apraxia or other movement disorders that each of these different neural regions makes its own
during the execution of self-paced voluntary movements distinct contribution to the representation of action.
[37 .]. Compared with a group of normal controls, two Progress in describing the unique contribution of each
patients with ideomotor apraxia showed electroencepha- region through the study of brain-damaged patients has
lograph patterns suggestive of decreased activity in left been limited by two factors: (a) the variability in the
parietal and medial frontal cortex. Unfortunately, no size, location and structures affected by the lesion and
structural images of the lesion site were available for (b) testing of a limited range of praxic functions, usually
these patients so it is not known whether the actual focusing solely on the performance of meaningful
lesion encompassed one or both of these sites. Analyzing gestures. Although meaningful actions are of greater
electroencephalographs in combination with structural interest to the apraxic patient who must be able to
imaging data would be a powerful tool permitting one to perform activities of daily living, relearning these actions
test the effects of a lesion in one brain region on activity would appear to require complex abilities in visuospatial
in another during tasks relevant to understanding analysis and the ability to reform action sequences from
apraxia. preserved movement abilities.

Treatment of apraxia Research on apraxia after stroke may continue to

It should be obvious that the presence of apraxia in the advance our current level of understanding of the neural
acute phase after stroke is an impediment to rehabilita- substrates of normal and impaired praxis. We make the
tive therapies aimed at improving ambulation and self- following recommendations for future studies: adequate
care, since the process of learning new motor sequences matching of groups with similar brain lesions in terms of
and skills usually depends heavily on imitative learning aetiology and onset; quantitative analyses of lesion
and on mental representation of an action. Moreover, in location and extent; assessment of the full range of
aphasic patients, the presence of apraxia prevents the praxic functions, including imitation of novel actions;
teaching of gestural communication as part of therapeu- separate assessment of different effectors, including
tic interventions. Although apraxia usually improves over mouth, face and foot; and exploration of patterns of
time following an acute lesion, spatiotemporal errors in neural activity associated with action representation in
imitation or tool use may persist [38]. The presence of normal controls and in apraxic patients.
limb apraxia, more than any other neuropsychological
disorder, correlates with the level of caregiver assistance Furthermore, it is now essential to begin the task of
required six months after a stroke [39], whereas the relating what is known about apraxia from the study of
absence of apraxia is a signicant predictor of return to patients, and about normal praxis from functional
76 Cerebrovascular disease

neuroimaging in normal controls, with the wealth of 3 Faglioni P, Basso A. Historical perspectives on neuroanatomical correlates of
limb apraxia. In: Neuropsychological studies of apraxia and related disorders.
literature describing the physiological mechanisms of Roy EA (editor). New York, NY: Elsevier Science Publishing Company, Inc.;
action knowledge and representation as observed in non- 1985. p. 344.
human primates. Studies of anatomic and functional 4 4 Roy EA, Square PA. Common considerations in the study of limb, verbal
connectivity will help to constrain our theories of which and oral apraxia. In: Neuropsychological studies of apraxia and related
disorders. Roy EA (editor). New York, NY: Elsevier Science Publishing
brain regions may work together systematically to Company, Inc.; 1985. pp. 111162.
subserve praxis. Lesions studies in non-human primates 5 Heilman KM, Maher LM, Greenwalk ML, Rothi LJ. Conceptual apraxia from
can be carefully controlled by the experimenter to allow lateralized lesions. Neurology 1997; 49:457464.
assessment of the effects of very focal unilateral or 6 Neiman MR, Duffy RJ, Belanger SA, Coelho CA. The assessment of limb
. apraxia: Relationship between performances on single- and multiple-object
bilateral damage on action representation. Finally, there tasks by left hemisphere damaged aphasic patients. Neuropsychol Rehabil
is a vast literature describing how single-cell recording 2000; 10:429448.
techniques have been used to dene the specic Study includes factor analysis of error types demonstrating equal loadings on
conceptual and motor execution factors for single- and multiple-object apraxia
response properties of neurons throughout the cortex tasks.
during reaching and grasping, during the presentation of 7 Rothi, LJ, Ochipa, C, Heilman, KM. A cognitive neuropsychological model of
graspable objects and during the observation of actions limb praxis. Cogn Neuropsychol 1991; 8:443458.
performed by others [4446]. 8 Cubelli R, Marchetti C, Boscolo G, Della Sala S. Cognition in action: testing a
. model of limb apraxia. Brain Cogn 2000; 44:144165.
Tests predictions derived from the model shown in Fig. 1 in a group of patients
The implications of such research for understanding with left hemisphere stroke.
apraxia are self-evident and indeed some physiologists 9 Goldenberg G. Matching and imitation of hand and finger postures in patients
have turned to the neurological research in patients in with damage in the left or right hemispheres. Neuropsychologia 1999;
37:559566.
interpreting their ndings. Unfortunately, the neuro-
10 Goldenberg G. Imitation and matching of hand and finger postures. Neuro-
logical literature contains few references to the physio- . . image 2001; 14:S132136.

logical literature as a framework for understanding the Review of recent research by the author and colleagues on the role of the left and
right hemispheres in praxis for novel actions. Emphasizes a role for the right
impairments observed in patients. A notable exception is hemisphere in visuospatial analysis of gestures.
the recent review by Leiguarda and Marsden [47], which
11 Buxbaum LJ, Giovannetti T, Libon D. The role of the dynamic body schema in
contains an impressive attempt at bringing together the . . praxis: evidence from primary progressive apraxia. Brain Cogn 2000; 44:166

literature from both camps. Clearly, the time has come to 191.
Detailed single-case analysis concludes with a qualitatively new model of praxis.
build a model of praxic functioning in which neurological The revised model provides a plausible account of praxis for novel actions,
and physiological data may be understood within a incorporating a dynamic body-schema representation that interacts with stored
knowledge of actions.
coherent framework. One approach would be to test the
12 Rizzolatti G, Fadiga L, Gallese V, Fogassi L. Premotor cortex and the
validity of neuropsychological models in terms of their recognition of motor actions. Cogn Brain Res 1996; 3:131141.
ability to map onto plausible physiological substrates. 13 Iacoboni M, Woods RP, Brass M, et al. Cortical mechanisms of human
We believe that a multidisciplinary approach will imitation. Science 1999; 286:25262528.
ultimately be most fruitful in terms of understanding 14 Nishitani N, Hari R. Temporal dynamics of cortical representation for action.
the nature, prognosis, and rehabilitation of apraxic Proc Natl Acad Sci U S A 2000; 97:913918.

impairments seen after stroke. 15 Donkervoort M, Dekker J, van den Ende E, et al. Prevalence of apraxia among
patients with a first left hemisphere stroke in rehabilitation centres and nursing
homes. Clin Rehabil 2000; 14:130136.
Acknowledgements 16 De Renzi E, Motti F, Nichelli P. Imitating gestures. A quantitative approach to
Supported by the Brain Mapping Medical Research Organization; the ideomotor apraxia. Arch Neurol 1980; 37:610.
Brain Mapping Support Foundation; the PiersonLovelace Foundation;
The Ahmanson Foundation; the Tamkin Foundation; the Jennifer Jones 17 Bizzozero I, Costato D, Della Sala S, et al. Upper and lower face apraxia: role of
. the right hemisphere. Brain 2000; 123:22132230.
Simon Foundation; the Capital Group Companies Charitable Founda-
tion; the Robson Family; the Northstar Fund and the National Center for Study of facial apraxia that demonstrates equal involvement of the left and right
Research Resources grants RR12169 and RR08655. hemispheres in the production of upper face gestures.

18 Roy EA, Heath M, Westwood D, et al. Task demands and limb apraxia in stroke.
. Brain Cogn 2000; 44:253279.
Large group (n = 99) study shows that a comparable proportion of patients with
References and recommended reading left hemisphere and right hemisphere strokes made apraxic errors during imitation
of transitive actions.
Papers of particular interest, published within the annual period of review, have
been highlighted as: 19 Weiss PH, Dohle C, Binkofski F, et al. Motor impairment in patients with
. of special interest parietal lesions: disturbances of meaningless arm movement sequences.
.. of outstanding interest Neuropsychologia 2001; 39:397405.
20 Goldenberg G. Defective imitation of gestures in patients with damage in the
left or right hemispheres. J Neurol Neurosurg Psychiatry 1996; 61:176180.
1 Leiguarda R. Limb apraxia: cortical or subcortical. Neuroimage 2001; 21 Goldenberg G, Hermsdorfer J, Laimgruber K. Imitation of gestures by
. 14:S137141. disconnected hemispheres. Neuropsychologia 2001; 39:14321443.
Review of the role of subcortical structures in praxis, including a critical discussion
of potential mechanisms of basal ganglia involvement. 22 Kertesz A, Ferro JM. Lesion size and location in ideomotor apraxia. Brain
1984; 107:921933.
2 Papagno C, Della Sala S, Basso A. Ideomotor apraxia without aphasia and 23 Schnider A, Hanlon RE, Alexander DN, Benson DF. Ideomotor apraxia:
aphasia without apraxia: the anatomical support for a double dissociation. J Behavioral dimensions and neuroanatomical basis. Brain Language 1997;
Neurol Neurosurg Psychiatry 1993; 56:286289. 58:125136.
 Apraxia after stroke Koski et al. 77

24 Halsband U, Schmitt J, Weyers M, et al. Recognition and imitation of 35 Kareken DA, Unverzagt F, Caldemeyer K et al. Functional brain imaging in
. pantomimed motor acts after unilateral parietal and premotor lesions: a apraxia. Arch Neurol 1998; 55:107113.
perspective on apraxia. Neuropsychologia 2001; 39:200216.
Study demonstrates that patients with circumscribed lesions to the left parietal 36 Peigneux P, Van Der Linden M, Andres-Benito P, et al. A neuropsychological
cortex are particularly apraxic for imitation of actions directed at their own body. and functional brain imaging study of visuo-imitative apraxia. Rev Neurol
2000; 156:459472.
25 Jancke L, Peters M, Himmelbach M, et al. fMRI study of bimanual
coordination. Neuropsychologia 2000; 38:164174. 37 Platz T, Kim IH, Pintschovius H, et al. Multimodal EEG analysis in man suggests
. impairment-specific changes in movement-related electric brain activity after
26 Haaland KY, Harrington DL, Knight RT. Neural representations of skilled stroke. Brain 2000; 123:24752490.
. movement. Brain 2000; 123:23062313. Novel approach to studying the electrophysiological correlates of self-paced
Study of stroke patients includes a quantitative analysis of the lesions on the basis meaningless movements in stroke patients with apraxia.
of neuroimaging data, suggesting importance of the intraparietal sulcus and the
middle frontal gyrus in goal-directed actions. 38 Maher LM, Ochipa C. Management and treatment of limb apraxia. In:
27 Jeannerod M. Neural simulation of action: a unifying mechanism for motor Apraxia: The neuropsychology of action. Rothi LJ, Heilman KM (editors).
cognition. Neuroimage 2001; 14:S103109. Hove, UK: Psychology Press; 1997. pp. 7592.

28 Hermsdorfer J, Goldenberg G, Wachsmuth C, et al. Cortical correlates of 39 Sundet K, Finset A, Reinvang I. Neuropsychological predictors in stroke
. gesture processing: Clues to the cerebral mechanisms underlying apraxia rehabilitation. J Clin Exp Neuropsychol 1988; 10:363379.
during the imitation of meaningless gestures. Neuroimage 2001; 14:149161.
PET study in normal subjects identifies cortical areas important for processing 40 Saeki S, Ogata H, Okubo T, et al. Return to work after stroke. A follow-up
observed gestures. study. Stroke 1995; 26:399401.

29 Damasio H, Grabowski TJ, Tranel D, et al. Neural correlates of naming 41 Smania N, Girardi F, Domenicali C, et al. The rehabilitation of limb apraxia: a
actions and of naming spatial relations. Neuroimage 2001; 13:10531064. study in left-brain-damaged patients. Arch Phys Med Rehabil 2000; 81:379
388.
30 Okada T, Tanaka S, Nakai T, et al. Naming of animals and tools: a functional
magnetic resonance imaging study of categorical differences in the human 42 van Heugten CM, Dekker J, Deelman BG, et al. Rehabilitation of stroke
brain areas commonly used for naming visually presented objects. Neurosci patients with apraxia: the role of additional cognitive and motor impairments.
Lett 2000; 296:3336. Disabil Rehabil 2000; 22:547554.
31 Grafton ST, Arbib MA, Fadiga L, Rizzolatti, G. Localization of grasp
43 van Heugten CM, Dekker J, Deelman BG, et al. Outcome of strategy training
representations in humans by positron emission tomography. 2. Observation
in stroke patients with apraxia: a phase II study. Clin Rehabil 1998; 12:294
compared with imagination. Exp Brain Res 1996; 112:103111.
303.
32 Buccino G, Binkofski F, Fink GR, et al. Action observation activates premotor
. and parietal areas in a somatotopic manner: an fMRI study. Eur J Neurosci 2001; 44 Fadiga L, Fogassi L, Gallese V, Rizzolatti G. Visuomotor neurons: ambiguity
13:400404. of the discharge or `motor' perception? Int J Psychophysiol 2000; 35:165
fMRI study in normal subjects shows somatotopic representations of observed 177.
actions.
45 Jellema T, Baker CI, Wicker B, Perrett DI. Neural representation for the
33 Peigneux P, Salmon E, van der Linden M, et al. The role of lateral perception of the intentionality of actions. Brain Cogn 2000; 44:280302.
occipitotemporal junction and area MT/V5 in the visual analysis of upper-limb
postures. Neuroimage 2000; 11:644655. 46 Umilta MA, Kohler E, Gallese V, et al. I know what you are doing. A
neurophysiological study. Neuron 2001; 31:155165.
34 Peigneux P, Salmon E, Garraux G, et al. Functional segregation of the
processing levels within a gesture-specific visual pathway. NeuroImage 47 Leiguarda RC, Marsden CD. Limb apraxias: higher-order disorders of
2001; 13:S927. sensorimotor integration. Brain 2000; 123:860879.