Exp Brain Res (1996) 110:265-278

9 Springer-Verlag 1996

Michel Desmurget 9 Claude Prablanc Mohammad Arzi 9Yves Rossetti 9Yves Paulignan Christian Urquizar

Integrated control of hand transport and orientation during prehension movements Received: 5 May 1995 / Accepted: 12 January 1996

At a descriptive level, prehension movements can be partitioned into three components ensuring, respectively, the transport of the arm to the vicinity of the target, the orientation of the hand according to object tilt, and the grasp itself. Several authors have suggested that this analytic description may be an operational principle for the organization of the motor system. This hypothesis, called "visuomotor channels hypothesis," is in particular supported by experiments showing a parallelism between the reach and grasp components of prehension movements. The purpose of the present study was to determine whether or not the generalization of the visuomotor channels hypothesis, from its initial form, restricted to the grasp and transport components, to its actual form, including the reach orientation and grasp components, may be well founded. Six subjects were required to reach and grasp cylindrical objects presented at a given location, with different orientations. During the movements, object orientation was either kept constant (unperturbed trials) or modified at movement onset (perturbed trials). Results showed that both wrist path (sequence of positions that the hand follows in space), and wrist trajectory (time sequence of the successive positions of the hand) were strongly affected by object orientation and by the occurrence of perturbations. These observations suggested strongly that arm transport and hand orientation were neither planned nor controlled independently. The significant linear regressions observed, with respect to the time, between arm displacement (integral of the magnitude of the velocity vector) and forearm rotation also supported this view. Interestingly, hand orientation was not implemented at only the distal level, demonstrating that all the redundant degrees of freedom available were used by the motor system to achieve the task. The final configuration reached by the arm was Abstract

M. Desmurget 9C. Prablanc (~) 9M. Arzi. Y. Rossetti Y. Paulignan. C. Urquizar Vision et Motricit6, Institut National de la Sant6 et de la Recherche M~dicaleUnit6 94, F-69500 Bron, France; Fax: +33-72-36-97-60, e-mail: [email protected]

very stable for a given final orientation of the object to grasp. In particular, when object tilt was suddenly modified at movement onset, the correction brought the upper limb into the same posture as that obtained when the object was initially presented along the final orientation reached after perturbation. Taken together, the results described in the present study suggest that arm transport and hand orientation do not constitute independent visuomotor channels. They also further suggest that prehension movements are programmed, from an initial configuration, to reach smoothly a final posture that corresponds to a given "location and orientation" as a whole. Key words Prehension 9Motor control 9Multijoint movement - Visual perturbation - Human

Introduction The problem of multijoint coordination, initially adressed by Bernstein (1967), remains very controversial, as demonstrated by the heterogeneity of the models proposed to account for the generation of complex actions involving simultaneously several musculoskeletal segments. One of the most heuristic theories proposed to explain the generation of goal-directed movement refers indisputably to the concept of "parallel visuomotor processing." This model implies that complex actions are segmented into functional modules implemented and controlled in parallel (Jeannerod 1981, 1992; Arbib 1981; Jeannerod and Biguer 1982; Arbib et al. 1985; Gentilucci et al. 1991; Paulignan et al. 1991a,b; Stelmach et al. 1994; Paulignan and Jeannerod, in press). According to this view, prehension movements directed to visual stimuli can be partitioned into three independent components ensuring respectively: the transport of the arm, the orientation of the hand, and the grasp (Arbib 1981; Jeannerod 1988, 1992). The concept of "parallel visuomotor channels" is supported by many neurophysiological studies. A first argument, suggesting the existence of a functional dissocia-

266 tion between the reach and grasp components of the prehension movements, is related to the observation that the corticospinal pathways responsible for the command of proximal and distal muscles are segregated (Brinkman and Kuypers 1972, Alstermark et al. 1981, Colebatch and Gandevia 1989, Kawashima et al. 1995). A second, similar argument arises from the functionnal organization of the inferior part of premotor area 6, which contains two types of neurons coding, respectively, the arm transport and the grasp formation (Gentilucci et al. 1988; Rizzolatti and al. 1988; Gentillucci and Rizzolatti 1990). The "transport neurons" respond specifically to a particular spatial location, and they appear to "drive" the hand toward a given point in space. Conversely, the "grasping neurons" respond distinctly to different type of grasping, and they appear to "control" the congruence between object size and the grasp pattern. Cells potentially related to different components of prehension movements are also present in the posterior part of the parietal cortex. Thus, Taira et al. 1990 (see also Gallese et al. 1994) have identified, in the posterior bank of the intraparietal sulcus (area 7a), some neurons selective to the type of object manipulated (grasping) or to the axis of orientation of the object (hand orientation). These neurons are distinct from other neurons, also located in area 7a, but activated during reaching movements (MacKay 1992). In addition to the above-mentioned neurophysiological studies, the parallel visuomotor channels hypothesis is also supported by a large number of psychophysical experiments. Thus, reaching for the same object presented at different locations in space influences the kinematics of the wrist (transport) without affecting the characteristics of the grasp (Jeannerod 1984; Gentilucci et al. 1991). Conversely, when objects of different size are presented at a given location, the modification of grip aperture does not influence the transport component of the movement, which remains stable (Wallace and Weeks 1988; Paulignan et al. 1991b). Constancy of arm transport can also be observed when the object to grasp is presented along different orientations in space (Lacquaniti and Soechting 1982; Stelmach et al. 1994). A last category of arguments in favor of a functional dissociation between reach and grasp can be drawn from data describing specific effects of cortical lesions or inactivations in man and animal (Jeannerod 1986, 1994; Chieffi et al. 1993; Gallese et al. 1994; Jeannerod et al. 1994). For example, specific deficits of hand preshaping were observed in monkeys after muscimol injections in the parietal cortex (anterior part of the lateral bank of the intraparietal sulcus, Gallese et al. 1994). In contrast with all of the above-mentioned experiments, many data seem to contradict the parallel visuomotor hypothesis by suggesting the existence of a functional coupling between the different components of prehension movements. Thus, selective perturbation of the size (Paulignan et al. 1991b; Castiello et al. 1992), the orientation (Stelmach et al. 1994), or the location (Paulignan et al. 1991a; Gentilucci et al. 1992) of the object to grasp influences all the components of the movement.

In the same way, when the arm is pulled back by a mechanical device during the movement, grip formation is strongly perturbed, suggesting that the hand aperture channel is afferented by the hand transport channel (Haggard and Wing 1991, 1995). Other arguments directly contradicting the visuomotor hypothesis can also be drawn from studies demonstrating both that movement time depends strongly on object size (Marteniuk and Mackenzie 1990) or texture (Fikes et al. 1994) and that the increase in the constraints acting on the arm during its displacement (fast movement, suppression of the vision of the arm, etc.) induces larger grip apertures (Wing and Fraser 1983; Wallace and Weeks 1988; Jakobson and Goodale 1991; Chieffi and Gentilucci 1993). Most of the data presented in the latter paragraph as arguments against the parallel visuomotor hypothesis are, however, not fully decisive. Indeed, their interpretation is far from being clear and unequivocal (for a detailed discussion of this point see Paulignan and Jeannerod, in press). For example, the absence of selective adaptation of one movement component (i.e., transport, orientation, or grasp) when the size, the orientation, or the location of the object is modified, can be explained by the existence of a coordinative mechanism linking temporally the different components of the movement (Jeannerod 1988, 1992; Paulignan et al. 1991a,b; Hoff and Arbib 1993). It can also be related to the use of discrete perturbations inasmuch as the successive presentation of two objects (Paulignan et al. 1991a,b; Castiello et al. 1992; Gentilucci et al. 1992; Stelmach et al. 1994) can induce the generation of two separate motor programs, as suggested by Gentilucci et al. (1992). At the same time, the link between object size and movement time can be due, in agreement with Fitts' law (Fitts 1954), to the variations of the area available for the grasp when object size is modified (Zaal and Bootsma 1993; Bootsma and al. 1994). Finally, it is possible that prior knowledge of the constraints imposed on the transport channel (no visual feedback, high velocity) will impose, prior to the movement, a change in the planned kinematic of the grasp (and conversely; Jeannerod 1988; Chieffi and Gentillucci 1993). The present study was designed to verify whether or not arm transport and hand orientation could be considered as "independent visuomotor channels." This hypothesis, which constitutes a generalization of the parallel model initially proposed by Jeannerod (1981; see also Arbib 1981) to account for the reach and grasp coordination during prehension movements implying precision grip, is principally supported by psychophysical studies. These studies show the lack of influence of the object orientation on the arm transport kinematics (Lacquaniti and Soechting 1982; Soechting 1984; Stelmach et al. 1994). Although indicative, this behavioral result seems, however, alone not sufficient to assert undoubtedly the independence of arm transport and hand orientation. In agreement with this view, it could be important to note that the studies supporting the generalization of the visuomotor channel hypothesis implied generally particular

267 e x p e r i m e n t a l situations (straight-line m o v e m e n t s ) , allowing a real p h y s i c a l i n d e p e n d e n c e b e t w e e n w r i s t d i s p l a c e m e n t and h a n d o r i e n t a t i o n (hand o r i e n t a t i o n r e m a i n s c o n s t a n t w h e n the w r i s t is d i s p l a c e d along a straight line). A s c l e a r l y shown b y a s i m p l e g e o m e t r i c a l analysis, this specific c a s e is far f r o m r e p r e s e n t a t i v e o f the g e n e r a l situation, in w h i c h the m o d i f i c a t i o n o f w r i s t p o s i t i o n i m plies a c o n c o m i t a n t m o d i f i c a t i o n o f a b s o l u t e o r i e n t a t i o n o f the h a n d in space. In other words, the t r a n s p o r t and o r i e n t a t i o n c o m p o n e n t s o f p r e h e n s i o n m o v e m e n t s are in m o s t cases (i.e., w h e n wrist d i s p l a c e m e n t is not restricte d to a straight line) not p h y s i c a l l y i n d e p e n d e n t , the final o r i e n t a t i o n o f the h a n d b e i n g c o n d i t i o n e d b y the spatial l o c a t i o n o f the w r i s t ( S o e c h t i n g and F l a n d e r s 1993). A n other t h e o r e t i c a l a r g u m e n t c h a l l e n g i n g the h y p o t h e s i s that a r m t r a n s p o r t and h a n d o r i e n t a t i o n constitute distinct and i n d e p e n d e n t v i s u o m o t o r c h a n n e l s refers to the existence o f b i - f u n c t i o n a l m u s c l e s . A s e m p h a s i z e d b y P a l a s t a n g a et al. 1994, m o s t o f the m u s c l e s s u p p o r t i n g a r m or f o r e a r m rotation are also flexors or extensors o f these two segments, w h i c h m a k e a total i n d e p e n d e n c e b e t w e e n the t r a n s p o r t and o r i e n t a t i o n c o m p o n e n t s o f the p r e h e n sion m o v e m e n t s t h e o r e t i c a l l y unlikely.

Materials and methods Recording technique Prehension movements were recorded, at a frequency of 200 Hz, by a SELSPOT II system, equipped with two cameras of 50 mm. A direct linear transform method (DLT) was used to reconstruct the three-dimensional coordinates of six infrared-emitting diodes (IREDs) placed on the right arm in the following positions (Fig. 1): (1) metacarpophalangeal joint of the index finger; (2) metacarpophalangeal joint of the auricular finger; (3) radial styloid; (4) cubital styloid, (5) above the cubital head of the elbow; (6) external extremity of the acromion. The center of gravity of IREDs 3 and 4 was used to reconstruct the transport component of the prehension movements. The distal IREDs (1, 2, 3, 4) were used to measure hand orientation. The shoulder and elbow IREDs (5, 6) were used to determine the participation of the proximal

IR6

~~

t Radio-cubital axis /

IR5

~ ~

,/

"-dR3/ / 9

/ / /

1

/ /

Metacarpo-phalangeal axis

IR2

/

Fig. 1 Schematic diagram illustrating the positioning of the six IREDs used in this experiment. Diodes were placed on the right arm as follows: (1) metacarpophalangeal joint of the index finger (IR1); (2) metacarpophalangeal joint of the auricular finger (IR2); (3) radial styloid (IR3); (4) cubital styloid (1174), (5) above the cubital head of the elbow (IRS); (6) external extremity of the acromion (IR6)

segments to hand orientation. Accuracy tests have indicated that the spatial precision of the recording system was below 2 mm. Subjects Six-right handed subjects (five men, one woman) from 22 to 48 years of age participated in this experiment9 All of them were naive with regard to the task and to the purpose of the experiment. Apparatus and procedure The apparatus and procedure used in this experiment are schematically represented in Fig. 2. The subject was seated in a totally dark room, in a dentist chair9 Their trunk was immobilized by a harness to prevent all displacement of the shoulder during the task. In front of the chair a servo-controlled torque motor supported the object to be grasped: a cylinder having a 400-g weight, a 5-cm diameter, and a 10-cm length. The object, which could be illuminated from inside through an electronic device, was located in the sagittal plane crossing the shoulder of the subject, at a distance corresponding to 80% of the upper-limb length. The motor allowed us to tilt the object within the frontoparallel plane along five different orientations: 60 ~ 40 ~ 20 ~ left (counterclockwise, +), 0 ~ (vertical), or 20 ~ right (clockwise, -). An electronic circuit detecting the onset of hand movement could control, via the motor, a very fast change in object orientation (40 ~ in less than 50 ms). A second electronic circuit located inside the object allowed the detection of the instant of contact between the hand and the object. At the beginning of each trial, the right arm of the subject rested on a tilted plane fixed on the side of the chair9 During the rest period, the subject was fixating on a central red LED placed in front of his body axis, at the same height and distance as that of the object. The arm was relaxed, with the hand at hip height and the right index finger in contact with the electrosensitive surface, allowing the detection of movement onset. The wrist was, like the forearm, in a neutral and comfortable position (flexion/extension, abduction/adduction, and prosupination angles, were all around 0~ After the cylinder was moved to one of its fixed orientations, a tone was given, and the central LED was turned off while the object was lit, indicating the subject had to grasp it "as quickly and accurately as possible" with a power grip. In this situation, "the object is held in a clamp formed by the partly flexed fingers and the palm, counterpressure being applied by the thumb lying more or less in the plane of the palm" (Napier 1956). The whole experiment was performed in the dark, and the upper limb was never visible, precluding any visual reafferences, except in the very last part of the movement, when the hand crossed the line of sight anchored on the lit object9 The experiment was divided into two sessions: 1. Session I consisted of a series of blocked trials (B). The object was then presented in a random order (ten repetitions) along five different orientations: 60 ~ (B60~ 40 ~ (B40~ 20 ~ (B20~ 0 ~ (BO~ ~ (B-20~ 2. Session II was identical to session I, except that the unperturbed trials (called here control trials, C) were randomly mixed with 20% of perturbed trials (P). In this latter case, the object initially presented at 20 ~ was suddenly and unexpectedly tilted at -20 ~ (P-20 ~ or 60 ~ (P60 ~ at movement onset. Each kind of unperturbed trial was repeated 16 times (16x5=80 movements), whereas each kind of perturbed trial was repeated ten times (10x2=20 movements). Before this second session, the general instructions from session I were repeated to the subject. A special emphasis was put on the necessity of keeping visual attention focused on the object, despite the apparent simplicity of the task. The subject was not informed of the possible occurrence of perturbations. Data analysis Position data were filtered at 10 Hz with a zero-phase finite impulse response filter (FIR) using 33 coefficients. The cutoff frequency was chosen after measuring the root-mean-square error

268

A Object

Illumination

Torque T,

motor

I, I

~and-Objeet Contact [ ~

Vlovement Onset r

Computer

Go signal

B Fixation point

T I

ON

[

OFF

oN

Step 2 (60~ Step 1 (20~ Perturbed trial

[

Perturbed trial

]

Unperturbed trial

[

-N

Step 1 (20~ Step 2 (-20~ Step 1 @

Position signal

J .....

Velocity

~ PV

TPV

[

~PA

Acceleration ,

,

t [

PD ~mD

I i

i

TM

r L

MT

i I

between the filtered and the nonfiltered signals for different possible frequencies ranging from 5 to 30 Hz. Ten hertz was found to be the best compromise for noise elimination and signal restitution. Hand velocity was computed from the filtered position signal, using a least square, second-order polynomial method (window_+4 points). The same method was applied for computing the acceleration. In order to verify whether arm transport and hand orientation could be considered as independent visuomotor

channels, two kinds of parameters were defined, spatial and kinematic. Spatial parameters In order to know whether wrist path was affected by object orientation, a first global parameter was computed, i.e., the total dis-

269 tance traveled by the wrist (TDT, the integral of the magnitude of the velocity vector). A second 3-D parameter, called linearity index (Lt), was also processed using a 3-D generalization of the index proposed by Atkeson and Hollerbach (1985). For each movement, the equation of the straight line joining the starting point and endpoint of the motion was computed. The largest deviation of arm path from that line was then determined (this intermediate variable was called wrist path absolute curvature, WPAC). LI was defined as the ratio between WPAC and the length of the line connecting the starting point and endpoint of the movement. The final 3-D spatial position of the elbow was also computed in order to determine whether the posture of the proximal segments was dependent or independent of object tilt. Elbow position was defined in a spherical frame of reference centered on the shoulder (x, sagittal axis; y, frontoparallel axis; z, vertical axis). Since the upper-limb length was constant (radius R, constant), only two variables were computed: elbow azimuth (EA; Fig. 3A) and elbow elevation (EE; Fig. 3A). EA was equal to 0 ~ when the upper arm was in the xz plane and to 90 ~ when the upper arm was in the yz plane. EE was equal to 0 ~ when the upper arm was horizontal and to - 9 0 ~ when the upper arm was oriented vertically downward. Note that final elbow position modifications were also assessed by a unique parameter (c~) representing the angle between the vertical axis (z) and the plane of the arm (plane containing the wrist, elbow, and shoulder; Fig. 3C). o~, which represents the rotation of the upper limb around the shoulder-object axis, was equal to 90 ~ when the upper limb was in the horizontal plane (xy) and to 180 ~ when the upper limb was downward in the sagittal plane (xz). Since the object was, in this experiment, tilted in the frontoparallel plane, the contribution of wrist movement (flexion-extension; abduction-adduction) to hand final orientation was weak, and the

9 Fig. 2 A Experimental setup. Subjects sat comfortably in a totally dark room. In front of them, a servo-controlled torque motor supported the cylinder to be grasped. At the beginning of each trial, the arm of the subject was relaxed on a tilted plane, the hand resting at hip height and the index finger being in contact with an electrosensitive surface, allowing the detection of movement onset, The object was located in the sagittal plane crossing the shoulder, at a distance corresponding to 80% of the whole upper-limb length. The orientation of the object could be modified before movement onset by rotating the motor in the frontoparallel plane (unperturbed trials). Object orientation could also be modified suddenly at movement onset through an electronic circuit detecting the onset of hand displacement (perturbed trials). These "online" modifications of object orientation were very fast (40 ~ in less than 50 ms). A central red LED, used as gaze-fixation point, was placed in front of the subject's body axis, at the same height and distance as those of the object. The time of contact between the hand and the object was detected automatically through an electronic device located inside the object. B Experimental procedure. The experiment was controlled on-line by a program run on an IBM 486 computer. During the rest period, the subject was instructed to fixate the central red LED. After the cylinder was moved in the dark to one of its fixed orientations, a tone was given, and the central LED was turned off while the object was lit, indicating the subject had to grasp it "as quickly and accurately as possible." Five orientations were examined: 60 ~ 40 ~ 20 ~ left (counterclockwise, positive), 0 ~ (vertical), and 20 ~ right (clockwise, negative). For the unperturbed trials, object orientation was not modified during the movement. For the perturbed trials, object orientation was suddenly modified at movement onset. Two types of perturbations were analyzed: 20~176 20~ ~. The whole experiment was performed in the dark. Arm movement was recorded at a frequency of 200 Hz, by a Selspot II system, equipped with two cameras of 50 mm. A direct linear transform method was used to reconstruct the three-dimensional coordinates of the infrared-emitting diodes placed on the right arm of the subject (MT movement time; PA, TPA peak and time-to-peak acceleration; PV, TPV peak and time-to-peak velocity; PV,, TPD peak and time-to-peak deceleration)

adjustment of hand orientation depended mainly on the radiocubiltal joint (Desmurget et al. 1994). Consequently, wrist motions were not analyzed in this work, and only the final forearm rotation angle (FR) was considered (Fig. 3B). This one was reconstructed through the computation of the angle separating the radiocubital vector (IREDs 3, 4) and the plane crossing, respectively, the shoulder (IRED 6), the elbow (IRED 5), and the wrist (center of gravity of IREDs 3 and 4). Internal rotations (pronations) were referred to as positive angles (0~ ~ and external rotation (supinations) as negative angles (0~176 Finally, a measure of the final orientation of the hand (FHO) with respect to the target was provided through the computation of the angle formed by the metacarpophalangeal vector (IREDs 1 and 2) and the axis of the object. This variable was computed one frame (5 ms) before hand/object contact, and it was used to quantify the grasping error due to the perturbations. Kinematic parameters In order to show whether arm transport was affected by hand orientation, wrist kinematics (kinematics of the center of gravity of the IREDs 3 and 4) were computed in a Cartesian frame of reference. All main temporal landmarks were determined (Fig. 2), i.e., movement time (MT, defined as being the duration between movement onset - release of the electrosensitive surface - and hand-object contact), peak and time4o-peak acceleration (PA and TPA), peak and time-to-peak velocity (PV and TPV), peak and time-topeak deceleration (PD and TPD). In order to test the influence of object tilt and the absence of specific expectation strategies during session II, a two-way ANOVA with repeated measures was performed on the unperturbed trials (blocked and control trials, one averaged value per subject and per angle), with object Orientation (p=5) and Session (q=2) being the repeated-measures factors. In order to compare the same number of trials, and since there were ten blocked trials, the analysis was preceded for each orientation by a random sorting of ten control trials among the 16 effectively performed. A second ANOVA with repeated measures was performed in order to test the effects of perturbations on prehension movements. Five conditions (repeated-measures factor) were considered, i.e., two perturbed conditions (P-20 ~ P60 ~ and three control conditions (C-20 ~ C20 ~ C60~ For the sake of simplicity, the repeated-measures factor was called the "orientation factor." Moreover, for the same reasons as those previously reported, and since there were ten perturbed trials, only control trials randomly selected in the first ANOVA were used here as the baseline for the study of the effects of perturbations. Threshold for statistical significance was set at o~=0.05. Duncan multiple-range tests were used for post hoc comparisons of the means. For each individual trial, the values of the distance traveled by the arm and the prosupination angle were calculated at each time. In order to determine whether arm transport and hand orientation covaried during the movement, simple linear regressions were computed between these two parameters for the extreme orientations of the target (-20 ~, 60~

Results

Unperturbed trials T a b l e s 1 a n d 2 s h o w the m e a n s and s t a n d a r d d e v i a t i o n s f o r the m o v e m e n t p a r a m e t e r s p r e v i o u s l y d e s c r i b e d ( b l o c k e d trials).

Analysis o f variance I n t e r a c t i o n s b e t w e e n o r i e n t a t i o n a n d s e s s i o n w e r e far f r o m s t a t i s t i c a l l y s i g n i f i c a n c e f o r all t h e k i n e m a t i c o r

270 Fig. 3A-C Schematic representation of the final angles computed in the present study (angles were computed at the time the hand touched the object). (D1 the sagittal plane crossing the shoulder, D2 the vertical plane containing the shoulder-elbow axis, D3 the plane containing the shoulder, S, the elbow, E, and the wrist, W, L1 horizontal line contained in D2 and crossing the shoulder). EA (i.e., the angle between D 1 and D2) characterized elbow azimuth (A); EE (i.e., the angle between L1 and the shoulder-elbow axis) characterized upper arm elevation (A); FR (i.e., the angle between D3 and the radiocubital axis) characterized forearm rotation (B), and ~ (i.e., the angle between D3 and the vertical) characterized the amount of rotation of the upper limb around the shoulder-object axis (C). EA=0 ~ when the upper arm was in the sagittal plane, and EA=90 ~ when the upper arm was in the frontoparallel plane. EE=0 ~ when the upper arm was horizontal, and EE=-90 ~ when the upper arm was oriented vertically downward, c~=90~ when the upper limb was in the horizontal plane, and c~=180~ when the upper limb was downward in the sagittal plane. FR=0 ~ when the radiocubital axis was in the upper-limb plane (D3). Internal rotations of the forearm (pronations) were referred to as positive angles (0~ ~ and external rotation (supinations) as negative angles (0~ ~

F R : Forearm Rotation

EA: Elbow Azimuth EE: Elbow Elevation

A

B Vertical Axis

Fronto-parallel Axis 0t: Upper Limb Rotation

C Table 1 Mean (and interindividual SDs) values of kinematic parameters during blocked trials (MT movement time, TPA time to peak acceleration, PA amplitude of peak acceleration, TPV time to

peak velocity, PV amplitude of peak velocity, TPD time to peak deceleration, PD amplitude of peak deceleration)

Object orientation

MT (ms)

TPA (ms)

PA (mm/s 2)

TPV (ms)

PV (mm/s 2)

TPD (ms)

PD (mm/s z)

60 ~ 40 ~ 20 ~ 0~ -20 ~

491 487 479 476 463

122 122 125 125 124

19271 19090 18033 17519 16160

209 212 217 220 222

2295 2260 2170 2104 2015

426 423 419 399 375

-17932 -18306 -17341 -16641 -15119

(80) (66) (71) (67) (52)

(22) (19) (24) (22) (20)

(5233) (5428) (4952) (4913) (3866)

spatial p a r a m e t e r s (highest F4,20=1.09, P>0.4). L i k e w i s e , n o n e o f the tested v a r i a b l e s was s i g n i f i c a n t l y affected b y the session factor (highest F1,5=1.06, P>0.3). In other words, c o n t r o l m o v e m e n t s p e r f o r m e d d u r i n g session II, as the u n p e r t u r b e d trials i n t e r m i x e d with the p e r t u r b e d one, w e r e i d e n t i c a l to b l o c k e d m o v e m e n t s p e r f o r m e d

(40) (35) (40) (36) (37)

(299) (312) (288) (273) (229)

(90) (75) (72) (61) (49)

(4662) (4547) (5063) (4812) (4229)

d u r i n g session I. C o n c e r n i n g the m a i n effect o f the orientation factor, the m o s t striking results w e r e as follows. Kinematic parameters. M o v e m e n t d u r a t i o n i n c r e a s e d significantly with object orientation (F4,20=3.90, P 60 ~

ooO -

A Unpertubed 20 ~

oo

o..y 0

e~

...~,,,,-~ DDO~~ /oooO

o~176 ."~

o U.p rtub d 20~

~ AAZ~ AZ~ ~ ' ~ ~

^~ .~ . . . . .

.. ~ "

ooo ..

9 Pertubed 20 ~ => -20 ~ ~J3

ee e

~o~ .."

oooo

ooO~ 0 El

o~176 o ~ ~ [][][]

oO 'oo~

0

o % ~ []o

200

250

550

SAGITTAL AXIS (mm) Fig. 4 Wrist path observed in the sagittal plane, for the unperturbed (white symbols) and the perturbed (black symbols) trials. Each individual curve was obtained after both temporal normalization and intersubjects averaging (n=6). For the unperturbed trials (only extreme and median orientations are represented), movement curvature appears clearly different for each orientation of the object. Moreover, the final location of the wrist is different according to object tilt (this difference is related to the modifications of elbow final location with respect to object orientation see text). These data suggest strongly that arm transport and hand orientation are not independently implemented. Wrist path corrections observed during perturbed trials bring the wrist into the same final position as those observed when the object is initially presented along the final orientation reached after perturbation (for instance a 20-60 ~ perturbation trial brings the wrist to the same position as the one reached when presenting an unperturbed 60 ~ trial). This result indicates that the transport and orientation components of the movement share information during the motion, i.e., that they are not independently controlled

cations o f m o v e m e n t temporal structure were also observed when object orientation was modified. In particular, D T ( D T = M T - T P V ) , measured either as a percentage or as an absolute value, decreased significantly with object tilt (lowest F4,20=11.66, P