Brain Advance Access originally published online on November 7, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Brain, Vol. 127, No. 1, 120-132, 2004
© 2004 Guarantors of Brain
doi: 10.1093/brain/awh006
Left and right hand recognition in upper limb amputees
1 Dipartimento di Psicologia, Università ‘La Sapienza’, 2 IRCCS Fondazione S.Lucia, Roma, Italy, 3 Institut de Sciences Cognitive, Bron, 4 Hôpital des Massues, Lyon, France and 5 Research Imaging Center, UTHSCSA, San Antonio, Texas, USA
Correspondence to: Angela Sirigu, Institute des Sciences Cognitives, 67, Blv. Pinel, 69675 Bron, France E-mail: sirigu{at}isc.cnrs.fr
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Previous research suggests a close similarity in brain activity between mental simulation of a movement and its real counterpart. To explore this similarity, we aimed to assess whether imagery is affected by the loss of a limb or of its motor skills. We examined the performance of 16 adult, upper limb amputees (and age-matched controls) in a left/right hand judgement task that implicitly requires motor imagery. The experimental group included subjects who had suffered the amputation of the dominant or the non-dominant limb. Although responding well above chance, amputees as a group were slower and less accurate than controls. Nevertheless, their response pattern was similar to that of controls, namely slower response times and more errors for stimuli depicting hands in unnatural orientations, i.e. postures difficult to reach with a real movement. Interestingly, for all stimuli, amputees’ performance was strongly affected by the side of limb loss: subjects who underwent amputation of their preferred limb made more errors and required greater latencies to respond as compared with amputees of the non-dominant limb. In a further analysis we observed that the habit of wearing an aesthetic prosthesis significantly interfered with the ability to judge the corresponding hand. Our data lead to three main conclusions: (i) loss of a single limb per se does not prevent motor imagery but it significantly enhances its difficulty; (ii) these subjects apparently perform the hand recognition task using a strategy in which they initially mentally simulate movements of their dominant limb; (iii) wearing a prosthesis, devoid of any motor function, seems to interfere with motor imagery, consistent with the view that only ‘tools’ can be incorporated in a dynamic body schema.
Key Words: hand recognition; amputation; motor imagery; hand preference; prosthesis
Abbreviations: ANOVA = analysis of variance; DP– = loss of dominant limb/not wearing prosthesis; DP+ = loss of dominant limb/wearing prosthesis; NDP– = loss of non-dominant limb/not wearing prosthesis; NDP+ = loss of non-dominant limb/wearing prosthesis; RT = response time
Received April 24, 2003. Revised July 31, 2003. Accepted August 1, 2003.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
In the past 20 years, varied and precise evidence has accumulated for a remarkable correspondence between properties of motor imagery and movement execution. Psychophysical studies using mental chronometry in normal subjects demonstrated how the time required to mentally simulate an action closely matches that needed to execute the corresponding motor act (Decety et al., 1989
; Jeannerod, 1995). More precisely, motor simulations seem to obey the same physical constraints (e.g. Fitt’s law on speed/accuracy trade-off) that apply to real movements (Sirigu et al., 1995, 1996). Such a parallelism is also found when the motor system is impaired: indeed, some patients suffering from somato-motor disorders show comparable perturbations in their motor imagery. When required to imagine movements of their hands, parkinsonian patients showed the same pattern of slowness and limb asymmetry that was observed during real motor execution (Dominey et al., 1995). Likewise, patients suffering from hemiparesis as a consequence of unilateral lesions of the motor cortex showed comparable slowness when executing and mentally simulating movements of their affected arm (Sirigu et al., 1995). Furthermore, recent neuroimaging studies demonstrate comparable activations during mental simulation and motor execution, suggesting the existence of a common neural substrate for the accompanying multimodal sensory-motor information processing that includes the parietal and premotor cortex, the basal ganglia and the cerebellum (Decety et al., 1994; Stephan et al., 1995; Grafton et al., 1996; Roth et al., 1996; Gerardin et al., 2000).
The tight similarity between imagery and motor control emerges also within tasks that implicitly activate motor imagery. For instance, in order to judge whether a stimulus presented in a picture is a left or a right hand, healthy subjects use a set of mental transformations that closely match the operations required for actual hand movements. Various studies strongly support this hypothesis (see a review in Parsons et al., 1998). In 1982, Sekiyama demonstrated that when subjects were asked to decide whether a right or a left hand was presented, their reaction times systematically varied reflecting hand-specific joint constraints. Namely, for each hand, greater reaction times were found for those positions that the arm and the hand could not easily reach with a real movement (Sekiyama, 1982). This response pattern revealed a preference for ‘manageable directions’ of actual movements, suggesting that subject’s judgments are likely to be based on a mental analogue that preserves kinaesthetic and/or proprioceptive information relative to the real movements. In another study, Parsons (1987a, b), using a task requiring the left/right judgement of a hand or a foot, confirmed that reaction times increase as a function of rotation angle of the stimulus. Interestingly, this effect is strongly influenced by the actual position of the subject’s body during the task, suggesting that subjects solve the task by mentally simulating their own body-part movement rather than by imagining a spatial transformation of a prototypical representation of a hand (i.e. a right hand in dorsal view, fingers pointing up). Thus, body representation appears to be the implicit functional base of motor activity also in the domain of mental simulation (Jeannerod, 1995; Jeannerod and Decety, 1995).
In agreement with this hypothesis, it has been shown that mentally simulated movements, like actual movements, probably respect the principle of control by the contralateral cerebral hemisphere. Parsons et al. (1998) investigated the mental representation of the hand in two split-brain patients. Subjects were required to judge whether a line drawing depicted either a right or a left hand: stimuli were presented in various orientations for 150 ms in either the left or right visual hemi-field. Patients’ performance was strongly affected by laterality of the stimulus: patients’ accuracy was equal to healthy controls when the hand presented on the screen was contralateral to the perceiving hemisphere, but did not rise above chance level for stimuli ipsilateral to the perceiving hemisphere. These results confirm that mental operations on body parts seem to depend on a contralateral cortical representation for each hand, in close analogy to overt motor control. A PET study on normal subjects with an analogue of Parsons’ experimental set, also confirmed that sensory-motor brain areas represent the mental simulation of shape and movement of the contralateral hand (Parsons et al., 1995).
To our knowledge, no study has directly addressed the question whether physical availability of the motor effector could be crucial to the solution of an implicit motor imagery task. Previous studies provide related and suggestive evidence. As discussed earlier, in a laterality judgement task, response times (RTs) to a visually presented hand shape seem to be strongly influenced by the current posture of the subject’s own hands (Parsons, 1994). This effect has been interpreted as a suggestion that subjects use their ‘first-person’ experience in order to mentally simulate the movement. This hypothesis was recently confirmed with an imagery task that directly compared ‘first-person’ and ‘third-person’ perspective (Sirigu and Duhamel, 2001), suggesting a close link between the mentally represented limb and its physical counterpart. Nevertheless, despite strong similarities, a complete overlap between the processes of action and mental simulation have been questioned. Neuroimaging studies suggest that when mentally simulating a movement, brain activity within the frontal lobe is more anterior with respect to overt motor execution, whereas activation within the parietal cortex shifts to more posterior regions (Gerardin et al., 2000). Furthermore, psychophysical studies demonstrate that patients who developed acute hemiplegia following a cerebral vascular accident retain the ability to use motor imagery of the paralysed limb in order to decide whether an overhand or an underhand grip were more appropriate to grasp a manipulandum (Johnson, 2000, Johnson et al., 2002). In other words, these patients seem to maintain the ability to mentally simulate movements of a body part they can no longer use.
The present research aimed to clarify these issues and to determine to what extent availability of a physical counterpart is crucial to motor imagery. To this purpose, we recorded the performance of a group of subjects who underwent amputation of the upper limb in a right/left hand-recognition task. We assumed that if implicit motor imagery does not require the presence of a physical counterpart, provided that a complete body representation had been established through prior experience, then amputees should be able to mentally simulate the movement of a body part even if that body part has been deleted. Alternatively, as several studies on parkinsonian and hemiplegic patients seem to suggest (Dominey et al., 1995; Sirigu et al., 1996), mental simulation of a movement might depend on the actual state of the body. If so, then the loss of a limb should interfere with the ability to recognize it, and accordingly amputees should show performances significantly different from that of normal controls. A secondary goal of the present study was to evaluate whether the habit of wearing a prosthetic limb has any effect on motor imagery. Tools are known to be functionally incorporated in the body schema (Iriki et al., 1996, 2001), therefore we were interested to evaluate whether and how an external object attached to one’s body in order to mimic a specific part, such as a prosthesis, affects motor imagery of the mimicked body part. With these two goals, we aimed to gain a better understanding of how the natural or artificial structural and functional integrity of one’s body sets the conditions for its motor-sensory imagery.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Participants
Sixteen subjects who had suffered amputation of the right or left upper limb were recruited at the Hôpital des Massues in Lyon (France). None of them had a previous history of neurological or psychiatric disorders. Seven among them (three women, four men; mean age 52.0 ± 16.1 years, range 28–74 years) had lost their dominant limb, whereas nine (all men; mean age 36.4 ± 15.5 years, range 21–63 years) suffered from the amputation of the non-dominant limb. According to the Edinburgh Inventory (Oldfield, 1971
), all but one (A8) were right-handed. All had normal or corrected to normal vision. The subjects’ main demographic and clinical data are summarized in Table 1 and in Supplementary data (available at Brain Online). A group of seven control subjects was recruited among relatives and medical staff (four women, three men; mean age 39.7 ± 15.3 years, range 24–64 years). All but one (C5) were right-handed according to the Edinburgh Inventory. All subjects had normal or corrected to normal vision. A pre-experimental analysis of variance (ANOVA) showed no significant differences between control subjects and the two groups of amputees with respect to mean age and educational level. As a control for traumatic limb loss, three subjects presenting congenital limb deletion of the left forearm (CD1–3, two women, one man, aged 22, 29 and 43 years) were also included in the study. To control for the loss of limb functionality rather than for its physical absence, two subjects having suffered a lesion of the right (PB1, right-handed woman, 36 years old) and left brachial plexus (PB2, right-handed man, 46 years old) were also tested. These subjects’ main demographic and clinical data are summarized in Supplementary data (available at Brain Online). In accordance with the local ethical committee, Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale (CCPPRB) Centre Léon Bérard, Lyon, which approved the study, all participants signed informed consent before volunteering for this study.
|
Stimuli
Stimuli were line drawings of both right and left hands, derived from Parsons (1987
a, b) and were presented as single images on a personal computer. Each drawing depicted one hand (approximately one-third of the size of the real hand) presented in one of four different viewpoints (see Fig. 1 for some examples). Viewpoints included two frontal postures (back and palm) and two side views (thumb side and pinkie side). For each viewpoint, hands were rotated through 12 different angles (in a 30° steps, from an arbitrary starting position with all fingers pointing up, corresponding to 0°/360° orientations as shown in Fig. 1). As judged by 12 naive subjects, six orientations corresponded to postures easily reached during normal movements (right hand, from 30° to 240° counter-clockwise; left hand, from 330° to 120° clockwise). The remaining six depicted postures requiring unnatural/uncomfortable movements in order to be matched (right hand: from 60° to 210° clockwise; left hand: from 300° to 150° counter-clockwise).
|
Procedure
Subjects sat comfortably in a dimly lit room and faced the screen of a portable computer located
30 cm from their frontal plane. They positioned their hands over their thighs and were instructed not to move them at all during the testing session. Subjects were required to look carefully at each drawing of a hand that appeared on the screen and to decide, as rapidly and accurately as possible, whether it was a right or left hand. The examiner started each trial by pressing a computer key. A fixation point appeared in the middle of the blank screen and remained visible for 200 ms. As soon as it disappeared, the image of one hand appeared in the same location; the drawing lasted on the screen until onset of the verbal response. Subjects were asked to answer by speaking aloud the words ‘droite’ or ‘gauche’ (French for ‘right’ and ‘left’, respectively). A voice-key microphone recorded response onset and terminated the trial by turning the screen blank. Both time and verbal response to each trial were recorded. RTs were computed as time elapsed between appearance of the line drawing on the screen and verbal onset of the response. RTs shorter than 300 ms and longer than 15000 ms were discarded from analyses. The identity of each verbal response was manually recorded by one experimenter. Two randomized sequences of 96 trials separated by a 15-min rest period were run in one testing session. Each sequence included 48 drawings of right hands and 48 of left hands presented in four different viewpoints (back, palm, thumb and pinkie side) and 12 orientations (six natural and six unnatural postures).
Data analysis
Accuracy for each participant was computed as the proportion of correct responses out of valid trials. This value was submitted to arcsine transformation and used for parametric analysis. Only RTs corresponding to valid trials were considered for analyses. RTs were submitted to logarithmic transformation in order to control for the effects of a skewed distribution and satisfy the conditions for parametric statistical test. For the purpose of the analysis, for each viewpoint, the 12 different orientations were grouped in two classes according to the difficulty of the real movement required to reach that posture (natural orientations, unnatural orientations), as judged by 12 naive subjects (see above). In order to assess whether our paradigm produced results congruent with previous reports (Sekiyama, 1982; Parsons, 1987a, b), two separate three-way ANOVAs for repeated measures (factors: hand presented – left, right; view – back, palm, thumb and pinkie side; orientation – natural, unnatural) were run on RTs and proportion of correct responses of control subjects only. Then, two separate four-way ANOVAs were run on proportion of correct responses and RTs to compare the control group with the amputees. The between-subjects factor was group (three levels: C, controls; D, subjects having lost their dominant limb; ND, subjects having lost their non-dominant limb); within-subjects factors were the same as in the previous analyses (hand x view x orientation). Finally, in order to evaluate the effect of wearing a prosthesis on motor imagery, amputees were grouped according to the loss of their dominant/non-dominant limb and to the habit of wearing/not wearing a prosthesis. Two separate four-way ANOVAs were run on proportion of correct responses and RTs, respectively. Between-subjects factor was group (five levels: C, control subjects; DP–, loss of dominant limb/not wearing prosthesis; DP+, loss of dominant limb/wearing prosthesis; NDP–, loss of non-dominant limb/not wearing prosthesis; NDP+, loss of non-dominant limb/wearing prosthesis) and the within-subjects factors were again hand x view x orientation.
Newman–Keuls Test was used for post hoc analysis of significant interactions.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
In an informal debriefing following the experimental session, most participants reported to have solved the task by mentally moving their own hand in order to reach the posture presented in the line drawing (69%). Several subjects described selecting what they considered the most plausible hand at first, and then switching to the other in case of error. Several amputees and subjects with congenital limb deletion, reported to have attempted to mentally simulate the movement of the present hand first. Fewer subjects, in addition to the aforementioned strategy, claimed to have based their response also on the thumb’s orientation relative to the wrist (31%).
At the time of testing, 12 amputees (A1, A2, A3, A4, A5, A6, A7, A10, A11, A13, A14, A16) and one subject with brachial plexus lesion (PB2) reported the presence of phantom sensations in their daily life. Interestingly, the phantom limb sensation was elicited or enhanced by the hand recognition task in nearly half of those subjects (A1, A3, A4, A14, A16), and in two of them (A1, A14) it was associated with phantom pain (see Table 2 for details on phantom sensations). In the latter two cases, subjects required several breaks during the task in order to overcome the painful sensation. A couple of subjects (A10, A11) reported actively rotating their phantom limb as well as their present limb during the experiment. Overall, subjective rating of task difficulty was higher among amputees than controls. Error rate was low in both controls (2.7 ± 2.2%) and amputees (12.6 ± 9.6%).
|
Control subjects
Results are summarized in Fig. 2 as the proportion of correct responses (left panel) and RTs (right panel) for stimuli depicting either a dominant (dark grey squares) or a non-dominant hand (light grey squares) in the different views and orientations.
|
Correct responses
As expected, a three-way ANOVA (hand x view x orientation) on proportion of correct responses of the control group revealed a main effect of stimulus-hand [F(1,6) = 7.524, P < 0.004; dominant hand, 0.98 ± 0.02; non-dominant hand, 0.97 ± 0.03] and orientation [F(1,6) = 5.736, P < 0.005; natural orientation, 0.99 ± 0.02; unnatural orientation, 0.97 ± 0.02]. Control subjects gave significantly more correct responses to images of their dominant hand. Moreover, natural orientations were more easily recognizable. The interaction between view and orientation reached statistical significance [F(3,18) = 4.800, P < 0.02]. Post hoc analysis showed that in the case of natural orientation, hands presented in pinkie-side views were significantly more difficult to recognize than hands presented in more commonly adopted postures, such as thumb (P < 0.05) or palm views (P < 0.05).
Response times
Analysis on RTs provided congruent results. An ANOVA showed main effects of all factors: stimulus–hand [F(1,6) = 11.724, P < 0.02; dominant hand, 1191.39 ± 124.94 ms; non-dominant hand, 1272.72 ± 172.12 ms]; view [F(3,18) = 11.331, P < 0.001; thumb side, 1102.57 ± 108.14 ms; back, 1150.45 ± 110.49 ms; palm, 1339.39 ± 162.08 ms; pinkie side, 1335.81 ± 56.98 ms] and orientation [F(1,6) = 19.540, P < 0.005; natural orientation, 1174.54 ± 158.27 ms; unnatural orientation, 1289.57 ± 128.08 ms]. Namely, RTs were faster when recognizing the dominant hand in the most natural orientations and views. The interaction between view and orientation was also significant [F(3,18) = 4.619, P < 0.02]. As previously described for correct responses, post hoc test showed that, for natural orientations, recognition of hands presented in pinkie-side views required significantly longer RTs than that of hands presented in more usual perspectives, such as thumb (P < 0.0003) or back views (P < 0.0005). Parsons’ study of movement time to these positions (for blocks of trials when the handedness of the stimuli was known by healthy subjects), showed that movement time to the pinkie side was longer than to the other views (Parsons, 1994). Thus, while visual unfamiliarity may be a contributing factor to the slower RTs (and somewhat less accurate responses), joint constraints and unfamiliar movement trajectories are likely to be primarily responsible for the greater mental simulation times and subsequent perceptual errors. Moreover, for unnatural orientations palm views required significantly longer RTs than thumb (P < 0.02) and back views (P < 0.04).
Summing up, in agreement with previous studies, control subjects produced slower and less accurate responses when judging a stimulus that depicted a hand in an unnatural orientation. Independent of this effect, control subjects were faster and more accurate judging stimuli portraying their dominant hand.
Control subjects versus amputees: effect of side of amputation
Correct responses
To examine whether the loss of the dominant limb versus the non-dominant limb had different effects on motor imagery and handedness judgment, amputees were grouped according to laterality of limb loss. A four-way ANOVA on proportion of correct responses showed main effects of group [F(2,20) = 4.304, P < 0.03; controls, 0.98 ± 0.02; D amputees, 0.87 ± 0.008; ND amputees, 0.90 ± 0.06], view [F(3,60) = 5.283, P < 0.003; thumb side: 0.93 ± 0.12 back: 0.92 ± 0.15, palm: 0.93 ± 0.10 and pinkie side: 0.88 ± 0.17] and orientation [F(1,20) = 15.598, P < 0.001; natural orientation, 0.94 ± 0.11; unnatural orientation, 0.89 ± 0.16]. Namely, subjects who had lost their dominant limb made significantly more errors in the task than controls (P < 0.02). On the contrary, amputees of the non-dominant limb did not differ from control subjects on overall measures. Unfamiliar postures, i.e. pinkie-side view, and unnatural orientations produced significantly more errors in all groups. Interaction between view and orientation was also significant [F(3,60) = 4.466, P < 0.007]. Post hoc test confirmed that fewer correct responses were reported by all participants for unnatural orientations (for all views: pinkie, 0.87 ± 0.20; thumb, 0.89 ± 0.14; back, 0.89 ± 0.17; palm, 0.90 ± 0.12). Accuracy increased for natural orientations, except for one condition: hands presented in pinkie side views (0.89 ± 0.14) were significantly more difficult to recognize than hands presented in more usual perspectives, such as thumb (0.97 ± 0.08, P < 0.0002), back (0.95 ± 0.12, P < 0.0005) and palm (0.95 ± 0.07, P < 0.003).
More interestingly, as suggested by the significant interaction between group, view and orientation [F(6,60) = 2.318, P < 0.05], a difference emerged among groups (see Fig. 3 for details, left panel). When line drawings depicted pinkie-side views and unnatural back views, subjects having suffered amputation of the dominant limb were significantly more impaired than those having lost their non-dominant limb (pinkie, P < 0.0003; back, P < 0.04) and control subjects (pinkie, P < 0.0002; back, P < 0.0002). No difference emerged between the latter two groups.
|
Response times
The same analysis on RTs showed main effects of all factors: group [F(2,20) = 4.253, P < 0.03; controls, 1232.05 ± 151.24 ms; D amputees, 2018.12 ± 315.42 ms; ND amputees, 1616.61 ± 181.32 ms], stimulus-hand [F(1,20) =15.383, P < 0.001; dominant hand, 1572.86 ± 712.13 ms, non-dominant hand, 1670.68 ± 675.73 ms], view [F(3,60) = 11.419, P < 0.00001; thumb side, 1476.09 ± 550.87; back, 1606.8 ± 869.30; palm, 1687.74 ± 665.42; pinkie side, 1713.70 ± 640.95] and orientation [F(1,20) = 26.516, P < 0.0001; natural orientation, 1501.35 ± 561.83; unnatural orientation, 1742.19 ± 789.77] (means and standard deviations are summarized in Fig. 3, right panel). More precisely, subjects having lost the dominant limb were significantly slower than controls (P < 0.02). Recognition of the non-dominant hand as well as of unfamiliar postures, i.e. pinkie-side view, and unnatural orientations required significantly longer RTs in all groups. This confirms the finding that the movement time to the natural orientations of the pinkie-side view is longer than to the natural orientations of the other stimulus views, in accordance with joint constraints and unfamiliar trajectories (Parsons, 1994
). Interaction between view and orientation was significant [F(3,60) = 10.109, P < 0.0001]. Post hoc test on the significant interaction confirmed that pinkie-side views were difficult to identify even in natural orientations and required significantly longer RTs (1696.85 ± 1730.56) than thumb-side (1371.38 ± 500.23, P < 0.0001), back (1426.07 ± 580.18, P < 0.0001) and palm views (1511.09 ± 555.25, P < 0.01). As for unnatural orientations, thumb-side views required shorter RTs (1580.79 ± 583.97) than the other perspectives (back, 1787.54 ± 1060.55, P < 0.004; palm, 1869.86 ± 732.03, P < 0.0001; pinkie 1730.56 ± 709.84, P < 0.005), although they were more difficult to recognize compared with the corresponding natural orientation (P < 0.0002). In summary, subjects having suffered amputation of the dominant limb made more errors and were significantly slower in solving the task compared with control subjects and amputees of the non-dominant limb. In particular, as can be seen in Fig. 3, amputees of the dominant limb were typically more impaired in recognition of unnatural or uncommon postures. Neither the age of subjects, nor elapsed time since amputation, were correlated with performance in the task.
Control subjects versus amputees: effect of wearing a prosthesis
Correct responses
To evaluate the role of the prosthesis in the mental simulation of body-part movements required by the task, amputees were grouped according to the habit of wearing (n = 9) or not wearing a prosthesis (n = 7, see Table 1). A four-way ANOVA on proportion of correct responses revealed a significant interaction between group and orientation [F(4,18) = 4.325, P < 0.02]. In addition, there were the expected effects of both view [F(3,54) = 5.634 P < 0.002; thumb 0.92 ± 0.07; back 0.91 ± 0.08; palm 0.92 ± 0.06; pinkie 0.88 ± 0.09] and orientation [F(1,18) = 20.719 P < 0.0003; natural 0.93 ± 0.06; unnatural 0.87 ± 0.09], as well as their interaction [view x orientation: F(3,54) = 3.301, P < 0.03]. As can be seen in Fig. 4 (left panel), and is confirmed by post hoc analysis, an interesting effect of wearing a prosthesis was found on subjects’ responses: the number of correct responses was significantly reduced in subjects wearing a prosthesis (filled symbols) compared with controls (grey area). This reduction was particularly striking for amputees of the dominant hand when judging on unnatural postures (P < 0.0002), but it emerged also in amputees of the non-dominant limb for both natural (P < 0.04) and unnatural postures (P < 0.04). Note that amputees of the non-dominant limb not wearing any prosthesis did not differ from controls.
|
Response times
Analysis of RTs showed a significant main effect of all factors: group [F(4,18) = 3.485, P < 0.03; controls, 1232.05 ± 151.24 ms; DP+ amputees, 2373.36 ± 494.2 ms; NDP+ amputees, 1732.78 ± 164.29; DP– amputees: 1544.46 ± 146.85 ms; NDP– amputees, 1471.39 ± 217.09 ms); stimulus hand [F(1,18) = 15.601, P < 0.001; dominant hand, 1621.05 ± 469.75 ms; non-dominant hand, 1720.56 ± 468.72 ms]; view [F(3,54) = 8.720, P < 0.0001; thumb side, 1667.66 ± 613.88; back, 1726.25 ± 429.37; palm, 1762.54 ± 447.82; pinkie side, 1580.72 ± 444.29 ms]; and orientation [F(1,18) = 30.184, P < 0.0001; natural orientation, 1542.16 ± 344.88; unnatural orientation, 1799.46 ± 540.88 ms]. Namely, subjects wearing a prosthesis were overall slower than controls in responding: this was particularly true for amputees of the dominant limb (filled symbols) compared with controls (grey area, P < 0.04, see Fig. 4).
Only the interaction between view and orientation was significant [F(3,54) = 8.771, P < 0.0001; thumb natural, 1371.38 ± 500.23 ms; thumb unnatural, 1580.79 ± 583.97 ms; back natural, 1426.07 ± 580.18 ms; back unnatural, 1787.54 ± 1060.55 ms; palm natural, 1511.09 ± 555.25 ms; palm unnatural, 1869.86 ± 722.03; pinkie natural, 1696.85 ± 571.24 ms; pinkie unnatural, 1730.56 ± 709.84 ms).
In conclusion, presence of a prosthetic limb significantly degraded performance in the present task, as shown by the increase of both errors and RTs. This effect was more pronounced for those subjects who lost their preferred limb and for responses to unnatural postures.
Control for limb loss
As a control for limb loss, we recruited three subjects presenting congenital limb deletion or having suffered from a lesion of the brachial plexus. Interestingly, those subjects, who never experienced presence of one upper limb, responded almost like control subjects (see Table 3 and Fig. 5) and did not differ from them as for proportion of correct hits. However, their RTs were overall slower than those of controls (see Fig. 5). Interestingly, these subjects did not show a tendency for longer RTs for unnatural postures for the deleted hand, but did show that tendency for the present hand. This suggests that the congenital absence of the limb precludes the ability to produce joint-constrained mental simulations for the deleted hand like those available for the present hand. At the same time, the high accuracy of the congenital deletion subjects may be a consequence of a strategy in which they judge the stimulus by always comparing it with their present hand: a mismatch in shape (a disconfirmation) implies that the stimulus must be the other hand. This interpretation is consistent with their subjective reports. Use of a successful disconfirmation decision strategy contrasts with the performance described in split brain patients (Parsons et al., 1998), who could judge with high accuracy only hands from the side of the body contralateral (and not ipsilateral) to the hemisphere viewing the stimuli. This difference indicates that each hemisphere of the split brain alone could not compare the hand for which it had accurate motor imagery to the viewed stimulus, in order to make successful use of a disconfirmation strategy.
|
|
An acquired peripheral loss of upper limb function, i.e. following lesion of the brachial plexus, strongly reduced performance in both the examined subjects (see Table 3 and Fig. 6), to a larger extent than in case of amputees wearing a prosthesis. The proportion of correct responses was quite small in both subjects, especially for items depicting the affected limb in unnatural postures. A consistent increase in RTs to the paralysed limb was also found in one subject (PB1).
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
It has been demonstrated in a variety of ways that a strong correlation exists between properties of motor imagery and actual motor behaviour (for reviews, see Decety and Ingvar, 1990
; Jeannerod and Decety, 1995; Crammond, 1997). However, to our knowledge the nature of the exact relationship between accuracy in mental simulation of a movement and actual state of the body is still poorly understood. Neuropsychological data on patients suffering from motor disabilities secondary to CNS damage suggest that motor imagery may take into account, as well as ignore, alterations affecting the motor system depending on the task (Dominey et al., 1995; Sirigu et al., 1996; Johnson, 2000, Johnson et al., 2002).
In the present study, we explored this issue by examining the case of peripheral modifications of the sensorimotor system. By contrast with previous studies, we assessed how motor imagery is affected by the physical absence of a motor equivalent in subjects who have no documented history of neurological impairment or CNS damage. Our subjects, who underwent amputation of either the dominant or non- dominant upper limb, were required to judge whether the image of a hand corresponded to a right or a left hand. It has been shown that this task implicitly activates motor imagery of the corresponding limb (Parsons, 1987a, b, 1994). Our study provides new information that can be summarized as follows. First, the loss of one limb does not prevent the ability to judge handedness, although it significantly increases task difficulty. Secondly, the loss of the dominant limb is a main source of perturbation, because the task is significantly more difficult for amputees who have lost their preferred hand. Thirdly, the everyday use of a prosthetic arm has a detrimental effect on the left/right judgement of a hand.
Our first finding confirms that, even if no explicit movement has to be performed, the handedness judgement task activates the motor system. The reported re-activation of extinguished phantom limb sensations in some amputees strongly support the claim that motor commands to the missing limb are elicited by our task and are still effective. This result is consistent with the report by Ramachandran and colleagues that false visual feed-back provided by a mirror image of the present limb can induce the feeling of motion in a previously extinguished phantom limb (Ramachandran, 1996). In addition, our data show that absence of the real effector reduces efficiency in motor mental simulation. There are alternative explanations of this observation.
The first possible explanation is that loss of a limb degrades the performance of the neural mechanisms that normally underlie movement and mental simulation. It is conceivable that the mental operation required by the task elicits a motor command that activates a predictive model of the final state, i.e. the posture that would eventually be achieved (Wolpert et al., 1999). In the absence of one limb, even if streams of motor commands can still be issued, no incoming information from the periphery is available. Thus, the feed-forward mechanism is no longer supported by information about either initial or final position of the missing hand. Without this mechanism of support for mentally simulating the movement of the missing limb, the subjects may resort to using an alternative strategy for stimuli representing that limb. If so, then the additional time required by amputees to solve the task may reflect use of a ‘visual’ imagery approach (i.e. a visual-spatial strategy rather than a motor-kinaesthetic one). Thus, in order to match the most difficult orientations, amputees may not use the strategy of mentally rotating their own hand to match the stimulus and choose instead to rotate the stimulus-hand as an object. However, on more familiar hand orientations, they may use a simple visual matching strategy. The latter possibility may be consistent with the reduction of errors for stimuli depicting familiar positions.
An alternative explanation is that amputees are slower in solving the task because of the change in body schema produced by the amputation. Indeed, limb loss may increase the difficulty of recognizing the missing hand because visual familiarity of the corresponding limb is no longer available. Moreover, this handedness judgement task requires manipulation of the internal representation of a body part and is known to activate brain areas devoted to somatic representation and body knowledge (Bonda et al., 1995; Parsons et al., 1995; Kawamichi et al., 1998). Since most of our amputees experienced phantom sensations in everyday life or during the task, it is possible that these subjects’ body image was still being adjusted to reflect the peripheral changes in their body, and perhaps this made it difficult to select the best strategy to solve the task. However, both of these body schema interpretations are contradicted by the findings on subjects with congenital limb deletion or brachial plexus lesion. Indeed, even when one limb is absent from birth and proprioceptive feed-back for movement has never been experienced, subjects are slower in judging the missing hand than the present one and show no prolonged RT effects related to joint-constrained motor imagery to awkward stimulus postures. This suggests that they are using alternative strategies, perhaps based on visual–spatial reasoning. Similarly, the two subjects suffering from brachial plexus lesion have lost the sensory-motor potential of their affected limb, still have its visual experience and a continuous feed-back of its presence, but are equally impaired in the task. Taken together, the performance of the experimental group emphasizes the effective role of availability of an intact motor potential. This effect may derive from the apparent requirement for motor commands to be issued, which thereby elicit failed or impoverished checks (on the basis of available proprioception) for feed-forward commands (Blakemore et al., 2002).
The second new finding here is that the left/right handedness judgments are slower and less accurate after loss of dominant limb than of non-dominant one. The longer RTs for dominant-limb amputees could not be attributed to an effect of the hand subjects used to respond, since no manual response was required. One alternative possibility is that the dominant-limb amputees used the visual–spatial strategy discussed earlier, which is more cognitively demanding (thus slower and more error prone). A second possible account is that amputees solved the task by mentally simulating movements of their preferred (missing) hand. Previous research (Sekiyama, 1982; Parsons, 1987a, b) has demonstrated that the most common strategy to judge handedness is a mental simulation of the real movement; namely, subjects mentally rotate their limb in order to match the orientation of the hand to be judged. It has been proposed that subjects implicitly choose by ‘guessing’ which hand to move first. After this automatic selection, a second explicit confirmatory phase would follow (Parsons, 1994; Parsons et al., 1995; Gentilucci et al., 1998). Our data are consistent with the possibility that, during the first phase, our subjects automatically select by default their dominant, preferred hand. If so, then when the subjects’ dominant limb is missing, they need to either switch to a visual imagery strategy, as previously suggested, or to use motor simulation by the present limb. Both of the latter operations would be likely to induce an increase in RTs and incorrect judgments. This hand preference for motor imagery is consistent with reports that motor asymmetries are present in the mental domain, affecting mental simulation tasks (Maruff et al., 1999) and movement attribution (Daprati and Sirigu, 2002). Moreover, in the present experiment, this advantage for stimuli depicting the dominant hand is supported by converging evidence: indeed, control subjects respond faster to stimuli depicting their dominant hand. Furthermore, subjects who suffered amputation of the non-dominant hand are not slower than controls.
A third possible explanation of the less efficient performance of the dominant-limb amputees is that loss and disuse of long-standing sensorimotor processes for the dominant limb may degrade in a general way the efficiency of both dominant and non-dominant upper limb motor behaviour and imagery. This is conceivable if the processes for the dominant limb, which are localized primarily in the dominant cerebral hemisphere, serve as the basis of abstract initial planning of all limbs motor behaviour. These three alternative accounts lead to predictions that could be readily tested in future studies.
The last finding we report is a surprising effect on imagery of wearing a prosthesis. In the nine amputees used to wearing a prosthesis daily, performance was significantly slower and less accurate than for controls and amputees not wearing a prosthesis. Most of these subjects used an aesthetic prosthesis, i.e. a rubber forearm and hand, or a mechanic device allowing a pinch grip by means of a shoulder movement that stretches a strip passing over the shoulder blade (see Table 1). Only two subjects were equipped with a myo-electric prosthesis allowing thumb opposition and wrist rotation by means of the contraction of the residual forearm muscles. Thus, this latter device affords movements that are bio-mechanic analogues of a few basic hand movements. Interestingly, performance of these two subjects was slightly better than the other amputees wearing prostheses.
Although conclusions must be drawn very cautiously from such a small sample, our data suggest that prostheses interfere with imagery when mental simulation of a movement is required. However, when the prosthesis can be used as an analogue of the missing hand, namely when it possesses a natural functionality, interference can be reduced. In this case, movement of the artificial hand re-establishes the possibility to update predictions issued by the motor outflow (Wolpert, 1997), by showing (on-line) the effect of the forearm muscles’ contraction. In contrast, an aesthetic prosthesis provides a visual feedback that emphasizes the ineffectiveness of motor commands, thus interfering with motor simulation. A similar effect is induced by the presence of a deafferented limb (i.e. following brachial plexus lesion). Indeed, our data show that performance of these subjects does not differ from that of amputees wearing an aesthetic prosthesis. This implies that a prosthesis can be incorporated in the body schema (and eventually improve mental simulation of a movement) when it works as a tool. This result would be in close agreement with studies on monkeys which showed that tools can become part of the body, being included in its representation (Iriki et al., 1996, 2001). These findings have important implications for prostheses applications, which are currently emerging via technology-intensive neural engineering approaches to assistive technologies.
|
Acknowledgements |
|---|
The authors wish to thank all the subjects who agreed to participate in the present study, as well as to the medical and technical staff of the Hôpital des Massues (Lyon, France) for their helpful assistance and Guillaume Fond for his help in collecting data. The research was supported by CNRS.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Blakemore SJ, Wolpert DM, Frith CD. Abnormalities in the awareness of action. Trends Cogn Sci 2002; 6: 237–42.[CrossRef][Web of Science][Medline]
Bonda E, Petrides M, Frey S, Evans A. Neural correlates of mental transformations of the body-in-space. Proc Natl Acad Sci USA 1995; 92: 11180–4.
Crammond DJ. Motor imagery: never in your wildest dream. Trends Neurosci 1997; 20: 54–7.[CrossRef][Web of Science][Medline]
Daprati E, Sirigu A. Laterality effects on motor awareness. Neuropsychologia 2002; 40: 1379–86.[CrossRef][Web of Science][Medline]
Decety J, Ingvar DH. Brain structures participating in mental simulation of motor behavior: a neuropsychological interpretation. Acta Psychol (Amst) 1990; 73: 13–34.
Decety J, Jeannerod M, Prablanc C. The timing of mentally represented actions. Behav Brain Res 1989; 34: 35–42.[Web of Science][Medline]
Decety J, Perani D, Jeannerod M, Bettinardi V, Tadary B, Woods R, et al. Mapping motor representations with positron emission tomography. Nature 1994; 371: 600–2.[CrossRef][Medline]
Dominey P, Decety J, Broussolle E, Chazot G, Jeannerod M. Motor imagery of a lateralized sequential task is asymmetrically slowed in hemi-Parkinson’s patients. Neuropsychologia 1995; 33: 727–41.[CrossRef][Web of Science][Medline]
Gentilucci M, Daprati E, Gangitano M. Right-handers and left-handers have different representations of their own hand. Brain Res Cogn Brain Res 1998; 6: 185–92.[CrossRef][Medline]
Gerardin E, Sirigu A, Lehericy S, Poline JB, Gaymard B, Marsault C, et al. Partially overlapping neural networks for real and imagined hand movements. Cereb Cortex 2000; 10: 1093–104.
Grafton ST, Arbib MA, Fadiga L, Rizzolatti G. Localization of grasp representations in humans by positron emission tomography. II. Observation compared with imagination. Exp Brain Res 1996; 112: 103–11.[Web of Science][Medline]
Iriki A, Tanaka M, Iwamura Y. Coding of modified body schema during tool use by macaque postcentral neurones. Neuroreport 1996; 7: 2325–30.[Web of Science][Medline]
Iriki A, Tanaka M, Obayashi S, Iwamura Y. Self-images in the video monitor coded by monkey intraparietal neurons. Neurosci Res 2001; 40: 163–73.[CrossRef][Web of Science][Medline]
Jeannerod M. Mental imagery in the motor context. Neuropsychologia 1995; 33: 1419–32.[CrossRef][Web of Science][Medline]
Jeannerod M, Decety J. Mental motor imagery: a window into the representational stages of action. Curr Opin Neurobiol 1995; 5: 727–32.[CrossRef][Web of Science][Medline]
Johnson SH. Imagining the impossible: intact motor representations in hemiplegics. Neuroreport 2000; 11: 729–32.[Web of Science][Medline]
Johnson SH, Sprehn G, Saykin AJ. Intact motor imagery in chronic upper limb hemiplegics: evidence for activity-independent action representations. J Cogn Neurosci 2002, 14: 841–52.[CrossRef][Web of Science][Medline]
Kawamichi H, Kikuchi Y, Endo H, Takeda T, Yoshizawa S. Temporal structure of implicit motor imagery in visual hand-shape discrimination as revealed by MEG. Neuroreport 1998; 9: 1127–32.[Web of Science][Medline]
Maruff P, Wilson PH, De Fazio J, Cerritelli B, Hedt A, Currie J. Asymmetries between dominant and non-dominant hands in real and imagined motor task performance. Neuropsychologia 1999; 37: 379–84.[CrossRef][Web of Science][Medline]
Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971; 9: 97–113.[CrossRef][Web of Science][Medline]
Parsons LM. Imagined spatial transformation of one’s body. J Exp Psychol Gen 1987a; 116: 172–91.[CrossRef][Web of Science][Medline]
Parsons LM. Imagined spatial transformations of one’s hands and feet. Cognit Psychol 1987b; 19: 178–241.[CrossRef][Web of Science][Medline]
Parsons LM. Temporal and kinematic properties of motor behavior reflected in mentally simulated action. J Exp Psychol Hum Percept Perform 1994; 20: 709–30.[CrossRef][Web of Science][Medline]
Parsons LM, Fox PT, Downs JH, Glass T, Hirsch TB, Martin CC, et al. Use of implicit motor imagery for visual shape discrimination as revealed by PET. Nature 1995; 375: 54–8.[CrossRef][Medline]
Parsons LM, Gabrieli JD, Phelps EA, Gazzaniga MS. Cerebrally lateralized mental representations of hand shape and movement. J Neurosci 1998; 18: 6539–48.
Ramachandran VS, Rogers-Ramachandran D. Synaesthesia in phantom limbs induced with mirrors. Proc R Soc Lond B Biol Sci 1996; 263: 377–86.[Medline]
Roth M, Decety J, Raybaudi M, Massarelli R, Delon-Martin C, Segebarth C, et al. Possible involvement of primary motor cortex in mentally simulated movement: a functional magnetic resonance imaging study. Neuroreport 1996; 7: 1280–84.[Web of Science][Medline]
Sekiyama K. Kinesthetic aspects of mental representations in the identification of left and right hands. Percept Psychophys 1982; 32: 89–95.[Web of Science][Medline]
Sirigu A, Duhamel JR. Motor and visual imagery as two complementary but neurally dissociable mental processes. J Cogn Neurosci 2001; 13: 910–19.[CrossRef][Web of Science][Medline]
Sirigu A, Cohen L, Duhamel JR, Pillon B, Dubois B, Agid Y, et al. Congruent unilateral impairments for real and imagined hand movements. Neuroreport 1995; 6: 997–1001.[Web of Science][Medline]
Sirigu A, Duhamel JR, Cohen L, Pillon B, Dubois B, Agid Y. The mental representation of hand movements after parietal cortex damage. Science 1996; 273: 1564–8.[Abstract]
Stephan KM, Fink GR, Passingham RE, Silbersweig D, Ceballos-Baumann AO, Frith CD, et al. Functional anatomy of the mental representation of upper extremity movements in healthy subjects. J Neurophysiol 1995; 73: 373–86.
Thobois S, Dominey PF, Decety PJ, Pollak PP, Gregoire MC, Le Bars PD, et al. Motor imagery in normal subjects and in asymmetrical Parkinson’s disease: a PET study. Neurology 2000; 55: 996–1002.
Wolpert DM. Computational approaches to motor control. Trends Cogn Sci 1997; 1: 209–16.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. L. Moseley and P. Brugger Interdependence of movement and anatomy persists when amputees learn a physiologically impossible movement of their phantom limb PNAS, November 3, 2009; 106(44): 18798 - 18802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Sharma, L. H. Simmons, P. S. Jones, D. J. Day, T. A. Carpenter, V. M. Pomeroy, E. A. Warburton, and J.-C. Baron Motor Imagery After Subcortical Stroke: A Functional Magnetic Resonance Imaging Study Stroke, April 1, 2009; 40(4): 1315 - 1324. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Dominici, E. Daprati, D. Nico, G. Cappellini, Y. P. Ivanenko, and F. Lacquaniti Changes in the Limb Kinematics and Walking-Distance Estimation After Shank Elongation: Evidence for a Locomotor Body Schema? J Neurophysiol, March 1, 2009; 101(3): 1419 - 1429. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Simmons, N. Sharma, J.-C. Baron, and V. M. Pomeroy Motor Imagery to Enhance Recovery After Subcortical Stroke: Who Might Benefit, Daily Dose, and Potential Effects Neurorehabil Neural Repair, September 1, 2008; 22(5): 458 - 467. [Abstract] [PDF]
|
|
|
|
 |
|
 G. L. Moseley Graded motor imagery for pathologic pain: A randomized controlled trial Neurology, December 26, 2006; 67(12): 2129 - 2134. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Fiorio, M. Tinazzi, and S. M. Aglioti Selective impairment of hand mental rotation in patients with focal hand dystonia Brain, January 1, 2006; 129(1): 47 - 54. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||