Neurops~choioyia, Vol. 29. No. 8. pp. 803-809. Printed in Great Britain.
1991
CQ28-3932;91 S3.00f0.00 d‘ 1991 Pergamon Press plc
NOTE A KINEMATIC
ANALYSIS OF REACHING AND GRASPING MOVEMENTS A PATIENT RECOVERING FROM OPTIC ATAXIA
L. S. *Department tDepartment
JAKOBSON,*
of Psychology, of Psychology,
Y.
M.
IN
ARCHIBALD,?’ D. P. CAREY* and M. A. GOODALE~$
University of Western Ontario, London, Ontario, Canada N6A 5C2; and Victoria Hospital, 375 South Street, London, Ontario, Canada N6A 4G5
(Received 11 October
1990; accepted
16 Apri/
1991)
Abstract-A detailed, kinematic analysis revealed subtle deficits in midline pointing and prehension in a patient showing good clinical signs of recovery from optic ataxia associated with bilateral parietooccipital damage. Relative to control subjects, the patient tended to misreach to the left with her right hand, and to the right with her left hand on a pointing task. While reach kinematics were otherwise normal in the pointing task, they were markedly disturbed in a prehension task, in which reaching and grasping movements must be integrated. In addition, difficulties in making fine postural adjustments to the hands were still evident 17 months post-injury. These findings suggest an important role for the posterior parietal lobes in programming goal-directed manual movements, and have implications for current theories of motor control and visual perception.
INTRODUCTION ATAXIArefers to a disorder of visually-guided hand movements not related to motor, somatosensory, visual field or visual acuity deficits. Although first reported in association with disorders of gaze control and visual attention [3, 111, the deficit can appear in isolation, following either unilateral or bilateral damage (e.g. [6]). Unilateral insult typically produces a disruption of reaching and grasping movements made by either hand in the visual field contralateral to the lesion [2,7,8,19,22]. In humans, optic ataxia is seen following circumscribed lesions of the posterior parietal lobe (especially areas 5 and 7 of the superior parietal lobule) 16, 191. One problem with much of the published work on optic ataxia is that, due to the technical difficulty of analysing complex movements, it has not been possible to study disturbances of prehension in detail. A few researchers [ 17,20, 211 have used video analysis to examine reaching and grasping following unilateral insult to the posterior parietal area. JEANNEKOD[17], for example, described grasping movements directed toward centrally presented targets by two such patients. In this region of space, movements are often described as being normal following unilateral damage. When these patients could see their movements, reaches made with the limb contralateral to the lesion were executed quite normally, although grip formation was disrupted. The removal of visual feedback, however, led to a complete loss of anticipatory grip formation, a more prolonged period of deceleration, and a systematic directional error in the reach toward the side of the lesion. Directional errors of this sort have also been described by others and suggest that vision may be required to keepegocentricand allocentriccoordinate systems in register in these patients (see [18], pp. 217-221, for a review of this argument). JEANNEROD’S1171 findings demonstrate how a detailed examination of movement kinematics may reveal subtle deficits not evident in standard clinical testing. In the present study, we carried out a kinematic analysis of reaching and grasping in a patient with hilaterul occipito-parietal damage who was showing only minima1 residual signs of optic ataxia in clinical assessment. The aim of the study was to contrast this patient’s ability to localize targets presented in center space in a pointing task with her ability to grasp objects in a prehension task. OPTIC
iReprint requests to: Dr M. A. Goodale, Ontario, Canada N6A X2.
Department
of Psychology,
803
University
of Western
Ontario,
London,
NOTE
804
METHOD Patient history Seventeen months prior to the present study, V.K. (a 68 year-old, right-handed woman) was admitted to hospital with cortical blindness secondary to congophilic angiopathy. A CT scan revealed bilateral haemorrhages in the occipital-parietal regions (see Fig. 1). One month later, the cortical blindness had resolved to a Balint’s syndrome [3], the component features of which include disordered gaze control, an impairment of visual attention and optic ataxia. There was no associated visual agnosia or anomia, and primary colours could be correctly identified. When the present tests were undertaken, V.K.‘s visual status was found to be much improved, although hesitancy and inaccuracy were still occasionally observed in her reaching and grasping performance. Central vision was wellpreserved both in formal perimetry testing and in testing to confrontation, but persisting difficulties were evident in her visual-scanning ability and in visuospatial/constructional skills. There was no evidence of manual apraxia or tactile sensory deficit in either hand. However, due to a slight residual weakness in the left hand, the present analysis focused on the performance of the right hand. Appuratus The positions of infra-red, light-emitting diodes (IREDs) attached to the right hand and arm were sampled at 100 Hz during two tasks with an opto-electronic recording system (WATSMART, Northern Digital inc., Waterloo, Canada). Two-dimensional digitized records were filtered off-line with a 7 Hz cutoff, using a second-order Butterworth filter, after which three-dimensional reach trajectories were reconstructed. Data analysis techniques and the accuracy of this system have been described elsewhere [13, 141. Procedure The patient’s performance on two tasks was compared to the performance of two control subjects (L.K. and B.S.), matched with the patient for age, sex and handedness. The first task involved pointing to light targets. To prevent subjects from learning a particular movement and simply repeating it, five different target locations were used. Thus, targets were presented 30 cm in front of the hand’s midline start position, or at positions 6 or 12 cm to the left or right ofthis central target. Only reaches to the central target were considered in the present analysis. The target lights were embedded in a wedge fastened to the table surface, and angled such that the greatest amount oflight from each target was directed toward the subjects’ eyes. During experimental sessions the target lights were concealed from view by black speaker cloth until independently illuminated, in a pre-specified sequence, by the experimenter. Subjects’ performance was assessed under full viewing conditions and when pointing movements had to be completed in the dark. In both viewing conditions, each target was presented once in each of six, randomly-ordered, five-trial sequences. Throughout testing, movements of an IRED marker attached to the tip of the index finger of the right hand were tracked and digitized with the WATSMART system, and stored on an IBM computer. Dependent variables considered in the present analysis included movement onset time, peak resultant velocity and the time at which it occurred, movement duration and terminal accuracy in the left/right dimension. The second task involved grasping an object which could be presented 20, 30 or 40 cm directly in front of the hand’s midline start position on a given trial. Three red, oblong blocks served as target objects (top surface dimensions 2 x 5 cm, 3 x 7.5 cm and 5 x 12.5 cm; all 2 cm in height). In preparation for a particular trial, the subject sat with eyes closed and with the index finger and thumb of the right hand touching one another and depressing the start key. After the subject opened her eyes, data collection began when the table surface was illuminated by an overhead fluorescent light. Subjects were instructed to pick up the target object as soon as it became visible. When the object was lifted a magnetic contact was broken, signalling the end of the collection period for that trial. In one block of trials, subjects’ movements were carried out under full viewing conditions, while in another block of trials the target object and the subject’s limb remained visible only until movement onset, when the overhead illumination was extinguished. In each viewing condition, four reaches were directed toward each target at each location in a random order, for a total of 36 experimental trials. IRED markers were attached to opposing sides of the nails of the right index finger and thumb, and opposite the styloid process of the ulna, proximal to the wrist ofthe right hand. Through three-dimensional reconstruction ofeach grasping movement, it was possible to provide a trialby-trial kinematic description of both the aiming movement of the arm (wrist marker: movement onset time, peak resultant velocity and the time at which this occurred, and movement duration) and of the timing and extent of the opening and closing of the hand (fingertip markers). Trials in which the object was dropped were repeated at the end of the currenf block of trials. The number ofsuch trials was negligible and did not differ between the patient and the controls. Only data from trials in which the target object was successfully retrieved were included in the analysis.
RESULTS
AND DISCUSSION
In the pointing task, the patient took somewhat longer to initiate her movements when she could see her moving limb than did the two control subjects. Nonethelesssheaccelerated to peak velocity quickly once the movement was underway. Moreover. V.K.‘s pointing movements to the central target were very comparable to those of control
NOTE
Fig. 1. CT scan from the patient (V.K.), showing bilateral lesions at the parieto-occipital junction. The size of the lesion in the right hemisphere (5 x 3 x 3 cm] was slightly larger than that in the left hemisphere (5 x 3 x 2 cm).
805
807
NOTE
subjects in terms of their peak resultant velocity and duration. What set the patient’s reaches apart from those of the controls was the fact that she made leftward terminal errors when pointing with the right hand. (Interestingly, rightward errors were evident when the patient pointed with the left hand, although this result is difficult to interpret due to the slight persisting weakness in the left limb.) These findings closely parallel published observations of opendifference. In our loop reaching in cases of unilateral posterior parietal damage (see [IS]), with one important patient, directional biases in the terminal accuracies ofpointing movements were apparent not only when the patient could not see her moving limb, but even under optimal viewing conditions (see Table 1). These data are consistent with JEANNEROD’S[17] conclusion that the parietal lobes may play an important role in keeping egocentric and allocentric maps in register.
Table 1. Mean values for a number of kinematic variables measured during pointing movements made with the right hand by the patient (V.K.) and two control subjects (L.K. and B.S.) (standard errors in brackets). Reaches were directed to a central target, located 30 cm in front of the hand’s start position Vision
V.K. L.K. B.S.
Movement onset (msec) 490 (70.0) 328 (14.5) 290 (18.3)
Time to peak velocity (msec) 165 (9.6) 347 (22.0) 322 (36.5)
Peak velocity (cmjsec) 89.3 (6.13) 75.6 (3.44) 70.2 (2.05)
Duration (msec) 935 (126.0) 797 (41.0) 987 (70.6)
Terminal error (cm)* -2.66 (0.16) 0.09 (0.05) -0.23 (0.24)
V.K. L.K. B.S.
Movement onset (msec) 430 (25.2) 423 (34.0) 548 (43.2)
Time to peak velocity (msec) 215 (17.1) 250 (47.2) 347 (46.5)
No vision Peak velocity (cmjsec) 66.9 (3.96) 75.4 (4.37) 61.1 (2.01)
Duration (msec) 1025 (188.2) 843 (89.7) 1122 (69.6)
Terminal error (cm)* - 1.95 (0.19) -0.43 (0.45) - I .39 (0.24)
S
S
*Measured
in the left/right
dimension
only. Leftward
errors
are negative,
rightward
errors
are positive.
Ifan equivalent directional bias was present in the grasping task, it was not sufficient to interfere with the patient’s ability to pick up small objects placed in front of her, even when she could not see her reaching limb or the target object during the approach phase. It is important to note, however, that the patient’s movement onset times were considerably longer in the prehension task than in the pointing task for targets an equivalent distance away (Table 2A). Once the movement was underway, however, the kinematics of the early portion of the reaching component were not greatly disturbed. Thus, as was the case with the control subjects, peak velocity was attained within the first 350 msec of the movement (Table 2B) and, although lower than the controls’, was scaled for movement amplitrtde (Table 2C). This was true regardless of whether or not vision of the moving limb was available after movement onset. Despite the relatively normal acceleration phase of the patient’s grasping movements, she showed an abnormally long period ofdeceleration. This effect is particularly clear when one compares movements of equivalent amplitude (i.e. 30 cm) in the pointing and prehension tasks directly. While the control subjects showed small increases in movement duration when the task changed from pointing to prehension, the patient’s movement duration times were essentially doubled by this change in task requirements (Table 2D). The maximum opening of the hand en route to the target typically occurs during the deceleration phase, when approximately 60% of total movement time has elapsed [14]. Because maximum grip aperture has been found in the past to be correlated with the size of the object to be grasped [ 15, 161, this measure is of some interest to researchers in the field of prehension. While the patient’s mean scores on this measure revealed weak evidence for scaling on the basis of object size, there was considerable trial-to-trial variability such that she often opened her hand wider in the presence of the small object than she did when the largest of the three objects was present. In fact, her overall maximum grip aperture was roughly 6&70% greater than that of the two normal controls (Table 2E), and she showed much more variability in timing the maximal opening of her hand (Table ZF). Frequent reposturing of the fingers was also evident during the deceleration phase of the kinematic profile as she closed in on the target object (Fig. 2). and this was exacerbated somewhat by the removal of visual feedback. These findings have some implications for models of human prehension. Researchers in the field often draw a distinction between two”components” of prehension movements: transport of the arm (which is often equated with an aiming movement) and formation of the grasp. In the present study, changing from a pointing task to a prehension task resulted in considerable disruption to the patient’s reaching movement, above and beyond the difficulty she experienced in making postural adjustments to her hand. Thus, relative to aiming movements, movement onset time and duration increased, and peak velocity decreased in the grasping task. These findings lend support to the argument made by CARNAHAN er al. [4] that the transport component of a grasping movement is
NOTE
808
Table 2. Changes in a number of kinematic variables as a function of task demands in the patient (V.K.) and two control subjects (L.K. and B.S.) (standard errors in brackets). The data are from the right hand. In all cases, only means from the condition in which visual feedback was continuously available are presented, although the same pattern was evident in the no feedback condition L.K.
B.S.
490 (70.0) 1302 (104.4)
328 (14.5) 789 (13.4)
290 (18.3) 584 (16.1)
315 (14.1)
339 (6.1)
222 (20.5)
V.K. A.
Mean movement onset (msec) (30 cm target distance)
B.
Mean time to peak velocity grasping task
C.
Mean peak velocity grasping task
D.
E.
F.
Task Pointing Grasping
(msec),
(cm/set),
Mean movement duration (30 cm target distance)
(msec)
Mean maximum grip aperture (mm), grasping task
Dist. to object 20 cm 30 cm 40 cm
33.2 (0.98) 42.9 (1.61) 54.9 (I .97)
73.3 (2.14) 96.3 (I .50) 117.7 (2.00)
77.2 (5.65) 90.0 (4.46) 102.2 (2.64)
Task Pointing Grasping
935 (126.0) 1846 (53.3)
797 (41 .O) 916 (19.4)
987 (70.6) I I I2 (34.2)
Object width 2cm 3 cm 5cm
146 (6.0) I48 (7.1) 162 (7.5)
73 (2.2) 84 (1.7) 100 (3.2)
86 (4.7) 99 (4.4) I I3 (2.5)
1018 (63.3)
604 (19.0)
651 (25.1)
Mean time to maximum grip aperture (msec), grasping task
0
500
1000
1500
2000 TIME
2500
3000
3500
/ 4000
(MS)
Fig. 2. Representative traces from single trials showing the change in grip aperture (the resultant distance between the index finger and thumb of the right hand) over time for the patient (V.K.) and two controls (L.K. and B.S.) when reaching to the same object located at the same distance from the start position. The flat portion at the beginning ofeach trace represents the subject’s movement onset time on that trial.
dependent on somewhat different sensorimotor networks than the corresponding component of a pointing movement. JEANNEROD[I 5,161 has suggested that the transport and grip formation components of a prehension movement involve separate visuomotor channels which are temporally coupled. Others 114) have argued that integrating the various component parts of a prehension movement (reaching, grip formation, rotation of the hand, etc.) to achieve fluid prehension requires considerable coordination between these systems, beyond mere temporal coupling. The
NATE
809
present finding of a disruption in the coordination of reaching and grasping following bilateral posterior parietal damage could reflect a disturbance in the mechanism normally responsible for integrating these components. The posterior parietal lobe is in an ideal position to act in this motor programming capacity, considering its extensive projections to premotor centers in the frontal lobe [l, 121, its role in “polysensory integration” [9] and its close association with the callosal fibers which connect the posterior areas of the two cerebral hemispheres [S]. Finally, the present results speak to the issue of whether or not perceiving objects is mediated by the same visual pathways as grasping them. Despite the fact that V.K. was unable to scale her grasp appropriately to the size of the target objects, she had no difficulty recognizing complex line drawings in formal tests of object recognition. Moreover, other work has shown that patients with optic ataxia are able to discriminate the orientation of lines presented in the very parts of the field in which they are unable to generate correctly oriented grasping movements [18]. In striking contrast to these patients, recent observations of visually guided grasping in a patient with visual form agnosia has revealed remarkably accurate calibration of grip formation with respect to the orientation and dimensions of objects, despite a profound inability to describe these same visual properties of objects or to match them by manual demonstration [IO]. Taken together, these results suggest that the mechanisms underlying the conscious perception of object qualities are quite separate from those controlling visually guided prehension. Acknowledgements-We would like to thank V.K. and the two control subjects for their cooperation. This research was supported by Medical Research Council ofcanada grant No. MA-7269 to M. A. Goodale. L. S. Jakobson and D. P. Carey are recipients of Medical Research Council Studentships.
REFERENCES 1. ANDERSON, R. A. Visual and eye movement functions of the posterior parietal cortex. Ann. Rev. Neurosci. 12, 377403, 1989. and unilateral optic ataxia associated with a left 2. AUEKBACH, S. H. and ALEXANDER, M. P. Pure agrdphia superior parietal lobule lesion. J. Neural., Neurosur(l. Psych&. 44,430-432, 1981. des “Schauens,” 3. BALINT, R. Seelenlahmung optlsche Ataxie, raumliche Storung der Aufmersamdeit. Monaf.s.schr$i Psychiat. Neural. 25, 57-8 1, 1909. versus pointing and the differential use of 4. CAKNAHAN, H., MAKTENIUK, R. G. and GOOUALE, M. A. Grasping visual feedback, submitted for publication. Ltd., London, 1953. 5. CKITCHLEY, M. The Parietal Lobes. Edward Arnold (Publishers) of hand movements under visual guidance. Neurology 29, 6. DAMASIO, A. R. and BENTON, A. L. Impairment 17@178, 1979. 7. FEKKO, J. M. Transient inaccuracy in reaching caused by a posterior parietal lobe lesion. J. Neural., Neurosurg. Psychiaf. 47, lOlf%lOl9, 1984. 8. FEKKO, J. M., BRAVO-MARQUES,J. M., CASTRO-CALDAS, A. and ANTUNES, L. Crossed optic ataxia: Possible role of the dorsal splenium. _I. Nrurol., Neurosurg. Psychiat. 46, 533-539, 1983. 9. FKEUND, H.-J. Clinical aspects of premotor function. Behav. Brain Res. 18, 187-191, 1985. dissociation between IO. GOOUALE, M. A., MILNEK, A. D., JAKOBSON, L. S. and CAREY, D. P. A neurological perceiving objects and grasping them. Nature 349, 154-l 56, 199 I II. HOLMES, G. Disturbances of visual orientation. Brit. J. Ophthal. 2, 449-468, 1918. 12. HYVAKINEN, J. Posterior parietal lobe of the primate brain. Physic>/. Rw. 62, 106&l 129, 1982. 13. JAKOBSON, L. S. and GOOUALE, M. A. Trajectories of reaches to prismatically-displaced targets: evidence for “automatic” visuomotor recalibration. E?cp. Brain Rex 78, 575-587, 1989. 14. JAKOBSON, L. S. and GOOUALE, M. A. Factors affecting higher-order movement planning: a kinematic analysis of human prehension. Eup. Bruin Rus., in press. 15. JEANNEKOU, M. Intersegmental coordination during reaching at natural visual objects. In Attention and Pr~/iwnancr IX, J. LONGand A. BAUUELEY(Editors), pp. 153-168. Erlbaum, Hillsdale, 1981. 16. JEANNEKOL).M. The timing of natural prehension movements. J. Mar. Behao. 16, 235-254, 1984. 17. JEANNEKOL),M. The formation of finger grip during prehension: A cortically mediated visuomotor pattern, Behur. Bruin Rrs. 19, 99-l 16, 1986. 18. JEANNEKOU, M. Thr Neurrrl and Beharioural Oryanization (I( Goal-Directed Morements. Clarendon Press,
Oxford.
1988.
19. LEVINE. D. N., KAUFMAN. K. J. and MOHK, J. P. Inaccurate
reaching associated with a superior parietal lobe tumor. Nuuro/o(/y 28, 556561, 1978. 20. PEKENIN, M.-T. and VIGHETTO, A. Optic ataxia: A specific disorder in visuomotor coordination. In Spatial/) Oriented Behwior, A. HEIN and M. JEAN~VEKOU (Editors), pp. 305-326. Springer-Verlag. New York, 1983. 21. PEKENIN. M.-T. and VIGHETTO. A. Optic ataxia: A specific disruption in visuomotor mechanisms. I. Different aspects of the deficit in reaching for objects. Bruin II 1, 643-674, 1988. 22. RONUOT. P.. RECONLM. J. DE and RIBADEAU DUMAS, J. L. Visuomotor ataxia. Bruin 100, 355-376, 1977.
1991
CQ28-3932;91 S3.00f0.00 d‘ 1991 Pergamon Press plc
NOTE A KINEMATIC
ANALYSIS OF REACHING AND GRASPING MOVEMENTS A PATIENT RECOVERING FROM OPTIC ATAXIA
L. S. *Department tDepartment
JAKOBSON,*
of Psychology, of Psychology,
Y.
M.
IN
ARCHIBALD,?’ D. P. CAREY* and M. A. GOODALE~$
University of Western Ontario, London, Ontario, Canada N6A 5C2; and Victoria Hospital, 375 South Street, London, Ontario, Canada N6A 4G5
(Received 11 October
1990; accepted
16 Apri/
1991)
Abstract-A detailed, kinematic analysis revealed subtle deficits in midline pointing and prehension in a patient showing good clinical signs of recovery from optic ataxia associated with bilateral parietooccipital damage. Relative to control subjects, the patient tended to misreach to the left with her right hand, and to the right with her left hand on a pointing task. While reach kinematics were otherwise normal in the pointing task, they were markedly disturbed in a prehension task, in which reaching and grasping movements must be integrated. In addition, difficulties in making fine postural adjustments to the hands were still evident 17 months post-injury. These findings suggest an important role for the posterior parietal lobes in programming goal-directed manual movements, and have implications for current theories of motor control and visual perception.
INTRODUCTION ATAXIArefers to a disorder of visually-guided hand movements not related to motor, somatosensory, visual field or visual acuity deficits. Although first reported in association with disorders of gaze control and visual attention [3, 111, the deficit can appear in isolation, following either unilateral or bilateral damage (e.g. [6]). Unilateral insult typically produces a disruption of reaching and grasping movements made by either hand in the visual field contralateral to the lesion [2,7,8,19,22]. In humans, optic ataxia is seen following circumscribed lesions of the posterior parietal lobe (especially areas 5 and 7 of the superior parietal lobule) 16, 191. One problem with much of the published work on optic ataxia is that, due to the technical difficulty of analysing complex movements, it has not been possible to study disturbances of prehension in detail. A few researchers [ 17,20, 211 have used video analysis to examine reaching and grasping following unilateral insult to the posterior parietal area. JEANNEKOD[17], for example, described grasping movements directed toward centrally presented targets by two such patients. In this region of space, movements are often described as being normal following unilateral damage. When these patients could see their movements, reaches made with the limb contralateral to the lesion were executed quite normally, although grip formation was disrupted. The removal of visual feedback, however, led to a complete loss of anticipatory grip formation, a more prolonged period of deceleration, and a systematic directional error in the reach toward the side of the lesion. Directional errors of this sort have also been described by others and suggest that vision may be required to keepegocentricand allocentriccoordinate systems in register in these patients (see [18], pp. 217-221, for a review of this argument). JEANNEROD’S1171 findings demonstrate how a detailed examination of movement kinematics may reveal subtle deficits not evident in standard clinical testing. In the present study, we carried out a kinematic analysis of reaching and grasping in a patient with hilaterul occipito-parietal damage who was showing only minima1 residual signs of optic ataxia in clinical assessment. The aim of the study was to contrast this patient’s ability to localize targets presented in center space in a pointing task with her ability to grasp objects in a prehension task. OPTIC
iReprint requests to: Dr M. A. Goodale, Ontario, Canada N6A X2.
Department
of Psychology,
803
University
of Western
Ontario,
London,
NOTE
804
METHOD Patient history Seventeen months prior to the present study, V.K. (a 68 year-old, right-handed woman) was admitted to hospital with cortical blindness secondary to congophilic angiopathy. A CT scan revealed bilateral haemorrhages in the occipital-parietal regions (see Fig. 1). One month later, the cortical blindness had resolved to a Balint’s syndrome [3], the component features of which include disordered gaze control, an impairment of visual attention and optic ataxia. There was no associated visual agnosia or anomia, and primary colours could be correctly identified. When the present tests were undertaken, V.K.‘s visual status was found to be much improved, although hesitancy and inaccuracy were still occasionally observed in her reaching and grasping performance. Central vision was wellpreserved both in formal perimetry testing and in testing to confrontation, but persisting difficulties were evident in her visual-scanning ability and in visuospatial/constructional skills. There was no evidence of manual apraxia or tactile sensory deficit in either hand. However, due to a slight residual weakness in the left hand, the present analysis focused on the performance of the right hand. Appuratus The positions of infra-red, light-emitting diodes (IREDs) attached to the right hand and arm were sampled at 100 Hz during two tasks with an opto-electronic recording system (WATSMART, Northern Digital inc., Waterloo, Canada). Two-dimensional digitized records were filtered off-line with a 7 Hz cutoff, using a second-order Butterworth filter, after which three-dimensional reach trajectories were reconstructed. Data analysis techniques and the accuracy of this system have been described elsewhere [13, 141. Procedure The patient’s performance on two tasks was compared to the performance of two control subjects (L.K. and B.S.), matched with the patient for age, sex and handedness. The first task involved pointing to light targets. To prevent subjects from learning a particular movement and simply repeating it, five different target locations were used. Thus, targets were presented 30 cm in front of the hand’s midline start position, or at positions 6 or 12 cm to the left or right ofthis central target. Only reaches to the central target were considered in the present analysis. The target lights were embedded in a wedge fastened to the table surface, and angled such that the greatest amount oflight from each target was directed toward the subjects’ eyes. During experimental sessions the target lights were concealed from view by black speaker cloth until independently illuminated, in a pre-specified sequence, by the experimenter. Subjects’ performance was assessed under full viewing conditions and when pointing movements had to be completed in the dark. In both viewing conditions, each target was presented once in each of six, randomly-ordered, five-trial sequences. Throughout testing, movements of an IRED marker attached to the tip of the index finger of the right hand were tracked and digitized with the WATSMART system, and stored on an IBM computer. Dependent variables considered in the present analysis included movement onset time, peak resultant velocity and the time at which it occurred, movement duration and terminal accuracy in the left/right dimension. The second task involved grasping an object which could be presented 20, 30 or 40 cm directly in front of the hand’s midline start position on a given trial. Three red, oblong blocks served as target objects (top surface dimensions 2 x 5 cm, 3 x 7.5 cm and 5 x 12.5 cm; all 2 cm in height). In preparation for a particular trial, the subject sat with eyes closed and with the index finger and thumb of the right hand touching one another and depressing the start key. After the subject opened her eyes, data collection began when the table surface was illuminated by an overhead fluorescent light. Subjects were instructed to pick up the target object as soon as it became visible. When the object was lifted a magnetic contact was broken, signalling the end of the collection period for that trial. In one block of trials, subjects’ movements were carried out under full viewing conditions, while in another block of trials the target object and the subject’s limb remained visible only until movement onset, when the overhead illumination was extinguished. In each viewing condition, four reaches were directed toward each target at each location in a random order, for a total of 36 experimental trials. IRED markers were attached to opposing sides of the nails of the right index finger and thumb, and opposite the styloid process of the ulna, proximal to the wrist ofthe right hand. Through three-dimensional reconstruction ofeach grasping movement, it was possible to provide a trialby-trial kinematic description of both the aiming movement of the arm (wrist marker: movement onset time, peak resultant velocity and the time at which this occurred, and movement duration) and of the timing and extent of the opening and closing of the hand (fingertip markers). Trials in which the object was dropped were repeated at the end of the currenf block of trials. The number ofsuch trials was negligible and did not differ between the patient and the controls. Only data from trials in which the target object was successfully retrieved were included in the analysis.
RESULTS
AND DISCUSSION
In the pointing task, the patient took somewhat longer to initiate her movements when she could see her moving limb than did the two control subjects. Nonethelesssheaccelerated to peak velocity quickly once the movement was underway. Moreover. V.K.‘s pointing movements to the central target were very comparable to those of control
NOTE
Fig. 1. CT scan from the patient (V.K.), showing bilateral lesions at the parieto-occipital junction. The size of the lesion in the right hemisphere (5 x 3 x 3 cm] was slightly larger than that in the left hemisphere (5 x 3 x 2 cm).
805
807
NOTE
subjects in terms of their peak resultant velocity and duration. What set the patient’s reaches apart from those of the controls was the fact that she made leftward terminal errors when pointing with the right hand. (Interestingly, rightward errors were evident when the patient pointed with the left hand, although this result is difficult to interpret due to the slight persisting weakness in the left limb.) These findings closely parallel published observations of opendifference. In our loop reaching in cases of unilateral posterior parietal damage (see [IS]), with one important patient, directional biases in the terminal accuracies ofpointing movements were apparent not only when the patient could not see her moving limb, but even under optimal viewing conditions (see Table 1). These data are consistent with JEANNEROD’S[17] conclusion that the parietal lobes may play an important role in keeping egocentric and allocentric maps in register.
Table 1. Mean values for a number of kinematic variables measured during pointing movements made with the right hand by the patient (V.K.) and two control subjects (L.K. and B.S.) (standard errors in brackets). Reaches were directed to a central target, located 30 cm in front of the hand’s start position Vision
V.K. L.K. B.S.
Movement onset (msec) 490 (70.0) 328 (14.5) 290 (18.3)
Time to peak velocity (msec) 165 (9.6) 347 (22.0) 322 (36.5)
Peak velocity (cmjsec) 89.3 (6.13) 75.6 (3.44) 70.2 (2.05)
Duration (msec) 935 (126.0) 797 (41.0) 987 (70.6)
Terminal error (cm)* -2.66 (0.16) 0.09 (0.05) -0.23 (0.24)
V.K. L.K. B.S.
Movement onset (msec) 430 (25.2) 423 (34.0) 548 (43.2)
Time to peak velocity (msec) 215 (17.1) 250 (47.2) 347 (46.5)
No vision Peak velocity (cmjsec) 66.9 (3.96) 75.4 (4.37) 61.1 (2.01)
Duration (msec) 1025 (188.2) 843 (89.7) 1122 (69.6)
Terminal error (cm)* - 1.95 (0.19) -0.43 (0.45) - I .39 (0.24)
S
S
*Measured
in the left/right
dimension
only. Leftward
errors
are negative,
rightward
errors
are positive.
Ifan equivalent directional bias was present in the grasping task, it was not sufficient to interfere with the patient’s ability to pick up small objects placed in front of her, even when she could not see her reaching limb or the target object during the approach phase. It is important to note, however, that the patient’s movement onset times were considerably longer in the prehension task than in the pointing task for targets an equivalent distance away (Table 2A). Once the movement was underway, however, the kinematics of the early portion of the reaching component were not greatly disturbed. Thus, as was the case with the control subjects, peak velocity was attained within the first 350 msec of the movement (Table 2B) and, although lower than the controls’, was scaled for movement amplitrtde (Table 2C). This was true regardless of whether or not vision of the moving limb was available after movement onset. Despite the relatively normal acceleration phase of the patient’s grasping movements, she showed an abnormally long period ofdeceleration. This effect is particularly clear when one compares movements of equivalent amplitude (i.e. 30 cm) in the pointing and prehension tasks directly. While the control subjects showed small increases in movement duration when the task changed from pointing to prehension, the patient’s movement duration times were essentially doubled by this change in task requirements (Table 2D). The maximum opening of the hand en route to the target typically occurs during the deceleration phase, when approximately 60% of total movement time has elapsed [14]. Because maximum grip aperture has been found in the past to be correlated with the size of the object to be grasped [ 15, 161, this measure is of some interest to researchers in the field of prehension. While the patient’s mean scores on this measure revealed weak evidence for scaling on the basis of object size, there was considerable trial-to-trial variability such that she often opened her hand wider in the presence of the small object than she did when the largest of the three objects was present. In fact, her overall maximum grip aperture was roughly 6&70% greater than that of the two normal controls (Table 2E), and she showed much more variability in timing the maximal opening of her hand (Table ZF). Frequent reposturing of the fingers was also evident during the deceleration phase of the kinematic profile as she closed in on the target object (Fig. 2). and this was exacerbated somewhat by the removal of visual feedback. These findings have some implications for models of human prehension. Researchers in the field often draw a distinction between two”components” of prehension movements: transport of the arm (which is often equated with an aiming movement) and formation of the grasp. In the present study, changing from a pointing task to a prehension task resulted in considerable disruption to the patient’s reaching movement, above and beyond the difficulty she experienced in making postural adjustments to her hand. Thus, relative to aiming movements, movement onset time and duration increased, and peak velocity decreased in the grasping task. These findings lend support to the argument made by CARNAHAN er al. [4] that the transport component of a grasping movement is
NOTE
808
Table 2. Changes in a number of kinematic variables as a function of task demands in the patient (V.K.) and two control subjects (L.K. and B.S.) (standard errors in brackets). The data are from the right hand. In all cases, only means from the condition in which visual feedback was continuously available are presented, although the same pattern was evident in the no feedback condition L.K.
B.S.
490 (70.0) 1302 (104.4)
328 (14.5) 789 (13.4)
290 (18.3) 584 (16.1)
315 (14.1)
339 (6.1)
222 (20.5)
V.K. A.
Mean movement onset (msec) (30 cm target distance)
B.
Mean time to peak velocity grasping task
C.
Mean peak velocity grasping task
D.
E.
F.
Task Pointing Grasping
(msec),
(cm/set),
Mean movement duration (30 cm target distance)
(msec)
Mean maximum grip aperture (mm), grasping task
Dist. to object 20 cm 30 cm 40 cm
33.2 (0.98) 42.9 (1.61) 54.9 (I .97)
73.3 (2.14) 96.3 (I .50) 117.7 (2.00)
77.2 (5.65) 90.0 (4.46) 102.2 (2.64)
Task Pointing Grasping
935 (126.0) 1846 (53.3)
797 (41 .O) 916 (19.4)
987 (70.6) I I I2 (34.2)
Object width 2cm 3 cm 5cm
146 (6.0) I48 (7.1) 162 (7.5)
73 (2.2) 84 (1.7) 100 (3.2)
86 (4.7) 99 (4.4) I I3 (2.5)
1018 (63.3)
604 (19.0)
651 (25.1)
Mean time to maximum grip aperture (msec), grasping task
0
500
1000
1500
2000 TIME
2500
3000
3500
/ 4000
(MS)
Fig. 2. Representative traces from single trials showing the change in grip aperture (the resultant distance between the index finger and thumb of the right hand) over time for the patient (V.K.) and two controls (L.K. and B.S.) when reaching to the same object located at the same distance from the start position. The flat portion at the beginning ofeach trace represents the subject’s movement onset time on that trial.
dependent on somewhat different sensorimotor networks than the corresponding component of a pointing movement. JEANNEROD[I 5,161 has suggested that the transport and grip formation components of a prehension movement involve separate visuomotor channels which are temporally coupled. Others 114) have argued that integrating the various component parts of a prehension movement (reaching, grip formation, rotation of the hand, etc.) to achieve fluid prehension requires considerable coordination between these systems, beyond mere temporal coupling. The
NATE
809
present finding of a disruption in the coordination of reaching and grasping following bilateral posterior parietal damage could reflect a disturbance in the mechanism normally responsible for integrating these components. The posterior parietal lobe is in an ideal position to act in this motor programming capacity, considering its extensive projections to premotor centers in the frontal lobe [l, 121, its role in “polysensory integration” [9] and its close association with the callosal fibers which connect the posterior areas of the two cerebral hemispheres [S]. Finally, the present results speak to the issue of whether or not perceiving objects is mediated by the same visual pathways as grasping them. Despite the fact that V.K. was unable to scale her grasp appropriately to the size of the target objects, she had no difficulty recognizing complex line drawings in formal tests of object recognition. Moreover, other work has shown that patients with optic ataxia are able to discriminate the orientation of lines presented in the very parts of the field in which they are unable to generate correctly oriented grasping movements [18]. In striking contrast to these patients, recent observations of visually guided grasping in a patient with visual form agnosia has revealed remarkably accurate calibration of grip formation with respect to the orientation and dimensions of objects, despite a profound inability to describe these same visual properties of objects or to match them by manual demonstration [IO]. Taken together, these results suggest that the mechanisms underlying the conscious perception of object qualities are quite separate from those controlling visually guided prehension. Acknowledgements-We would like to thank V.K. and the two control subjects for their cooperation. This research was supported by Medical Research Council ofcanada grant No. MA-7269 to M. A. Goodale. L. S. Jakobson and D. P. Carey are recipients of Medical Research Council Studentships.
REFERENCES 1. ANDERSON, R. A. Visual and eye movement functions of the posterior parietal cortex. Ann. Rev. Neurosci. 12, 377403, 1989. and unilateral optic ataxia associated with a left 2. AUEKBACH, S. H. and ALEXANDER, M. P. Pure agrdphia superior parietal lobule lesion. J. Neural., Neurosur(l. Psych&. 44,430-432, 1981. des “Schauens,” 3. BALINT, R. Seelenlahmung optlsche Ataxie, raumliche Storung der Aufmersamdeit. Monaf.s.schr$i Psychiat. Neural. 25, 57-8 1, 1909. versus pointing and the differential use of 4. CAKNAHAN, H., MAKTENIUK, R. G. and GOOUALE, M. A. Grasping visual feedback, submitted for publication. Ltd., London, 1953. 5. CKITCHLEY, M. The Parietal Lobes. Edward Arnold (Publishers) of hand movements under visual guidance. Neurology 29, 6. DAMASIO, A. R. and BENTON, A. L. Impairment 17@178, 1979. 7. FEKKO, J. M. Transient inaccuracy in reaching caused by a posterior parietal lobe lesion. J. Neural., Neurosurg. Psychiaf. 47, lOlf%lOl9, 1984. 8. FEKKO, J. M., BRAVO-MARQUES,J. M., CASTRO-CALDAS, A. and ANTUNES, L. Crossed optic ataxia: Possible role of the dorsal splenium. _I. Nrurol., Neurosurg. Psychiat. 46, 533-539, 1983. 9. FKEUND, H.-J. Clinical aspects of premotor function. Behav. Brain Res. 18, 187-191, 1985. dissociation between IO. GOOUALE, M. A., MILNEK, A. D., JAKOBSON, L. S. and CAREY, D. P. A neurological perceiving objects and grasping them. Nature 349, 154-l 56, 199 I II. HOLMES, G. Disturbances of visual orientation. Brit. J. Ophthal. 2, 449-468, 1918. 12. HYVAKINEN, J. Posterior parietal lobe of the primate brain. Physic>/. Rw. 62, 106&l 129, 1982. 13. JAKOBSON, L. S. and GOOUALE, M. A. Trajectories of reaches to prismatically-displaced targets: evidence for “automatic” visuomotor recalibration. E?cp. Brain Rex 78, 575-587, 1989. 14. JAKOBSON, L. S. and GOOUALE, M. A. Factors affecting higher-order movement planning: a kinematic analysis of human prehension. Eup. Bruin Rus., in press. 15. JEANNEKOU, M. Intersegmental coordination during reaching at natural visual objects. In Attention and Pr~/iwnancr IX, J. LONGand A. BAUUELEY(Editors), pp. 153-168. Erlbaum, Hillsdale, 1981. 16. JEANNEKOL).M. The timing of natural prehension movements. J. Mar. Behao. 16, 235-254, 1984. 17. JEANNEKOL),M. The formation of finger grip during prehension: A cortically mediated visuomotor pattern, Behur. Bruin Rrs. 19, 99-l 16, 1986. 18. JEANNEKOU, M. Thr Neurrrl and Beharioural Oryanization (I( Goal-Directed Morements. Clarendon Press,
Oxford.
1988.
19. LEVINE. D. N., KAUFMAN. K. J. and MOHK, J. P. Inaccurate
reaching associated with a superior parietal lobe tumor. Nuuro/o(/y 28, 556561, 1978. 20. PEKENIN, M.-T. and VIGHETTO, A. Optic ataxia: A specific disorder in visuomotor coordination. In Spatial/) Oriented Behwior, A. HEIN and M. JEAN~VEKOU (Editors), pp. 305-326. Springer-Verlag. New York, 1983. 21. PEKENIN. M.-T. and VIGHETTO. A. Optic ataxia: A specific disruption in visuomotor mechanisms. I. Different aspects of the deficit in reaching for objects. Bruin II 1, 643-674, 1988. 22. RONUOT. P.. RECONLM. J. DE and RIBADEAU DUMAS, J. L. Visuomotor ataxia. Bruin 100, 355-376, 1977.
