Clinical Rehabilitation 2004; 18: 4859

Neurophysiological and behavioural adaptations to a bilateral training intervention in individuals following stroke
Gwyn N Lewis and Winston D Byblow Human Motor Control Laboratory, Department of Sport and Exercise Science, University of Auckland, New Zealand Received 23rd February 2003; returned for revisions 10th July 2003; revised manuscript accepted 5th August 2003.

Objective: To examine the neurophysiological and behavioural adaptations to a bilateral training protocol in individuals with hemiparesis due to stroke. Design: Subjects performed daily practice sessions of three speci c upper limb tasks over a four-week period. Performance of the tasks was recorded on video for later analysis over the duration of the study. For the rst week of the training period the tasks were practised with the impaired upper limb. Only then, with a random start day between 8 and 13, the tasks were practised using a bilateral training protocol. Functional assessments of the affected upper limb and neurophysiological investigations of contra- and ipsilateral pathways from the affected and nonaffected cortical hemispheres were completed prior to the start of the intervention, at the end of the unimanual practice period, and at the conclusion of the bimanual practice period. Subjects: Six individuals between one month and four years post stroke. Interventions : Unimanual and bimanual training protocols. Main measures : Task performance with the affected limb only; motor impairment of the affected upper limb; ipsi- and contralateral corticospinal pathway excitability. Results: Baseline task performance varied markedly between subjects. Most individuals demonstrated little, if any, additional bene cial effect of bimanual practice on task performance. Results of neurophysiological investigations were inconsistent between subjects. Conclusions: Short-term bilateral training following unilateral training may have limited effectiveness in enhancing upper limb motor performance in acute and chronic individuals poststroke. The neural mechanisms associated with bilateral activation therapies remain uncertain.

Address for correspondence: Gwyn N Lewis, Sensory Motor Performance Program, Rehabilitation Institute of Chicago, 345 E Superior St, Chicago, IL 60611, USA. e-mail: g-lewis3@northwestern.edu Arnold 2004 10.1191/0269215504cr701oa

Bilateral training intervention post stroke Introduction In recent years the development of new rehabilitation therapies has demonstrated that signi cant progressions in movement ability are achievable in chronic stroke patients many months or even years after the initial event.1,2 One of these recent approaches is the use of bilateral activation therapies. A number of studies have examined the short- and long-term alterations in movement pro les of the affected limb of hemiplegic individuals in response to simultaneous bilateral movement protocols. Several authors have demonstrated a bene cial effect of concurrent, isokinematic bilateral activation on movement parameters of the affected side in individuals with congenital hemiparesis and following stroke.3,4 However, others have shown that, during bimanual activation, the impaired limb appears to constrain the movements of the nonimpaired limb, leaving the performance of the affected side unchanged.5,6 More promising results in terms of functional ability of the affected limb have been seen with longer term bilateral training interventions in individuals poststroke. Mudie and Matyas1,7 demonstrated an enhancement in task performance after a three-week period of bilateral isokinematic therapy (BIT). In these studies, simultaneous bilateral practice of motor tasks with both upper limbs was seen to elicit improvements in unilateral task performance with the affected limb only. The extent of motor improvement with bilateral training was greater than that seen in response to a similar period of unilateral practice, or practice sessions in which the nonaffected limb assisted the affected limb in carrying out the tasks. Whitall and colleagues8 also reported signi cant increases in strength, range of motion, and standardized motor assessment scales following six weeks of a simultaneous bilateral arm training protocol, while Cauraugh and Kim9 reported enhanced motor control after electromyogram-triggered neuromuscular stimulation coupled with bilateral movement. The above studies provide evidence that bilateral training techniques have the potential to elicit endurable improvements in movement ability in chronic stroke patients, and that these effects are obvious in unilateral performance and

49

in general motor function. There are several possibilities as to the neural mechanisms and pathways that are targeted by bilateral activation interventions. These include an interaction in the affected hemisphere between spared cells of ipsilateral pathways and spared cells of the crossed corticospinal pathways; a facilitation of ipsilateral pathways from the nonaffected hemisphere; and the involvement of indirect pathways from the affected hemisphere, such as the rubrospinal or reticulospinal.1 Currently, there is little neurophysiological evidence to indicate which, if any, of these pathways may be activated in response to bilateral training techniques. The aim of the present study was to examine the neurophysiological and behavioural adaptations to a four-week bilateral training protocol in individuals following stroke. As well as measures of task performance and motor impairment of the upper limb over the course of the training protocol, neurophysiological parameters were also investigated using transcranial magnetic stimulation (TMS). Methods Participants Six individuals who had suffered a mono-hemispheric stroke participated in the study (Table 1). Criteria for inclusion in the study included: evidence of a mono-hemispheric lesion documented by computerized tomography (CT) or magnetic resonance imaging (MRI); rst-ever stroke; absence of any cognitive impairments or other neurological disorders; impairment of upper limb function relative to the noninvolved side; and no contraindications to TMS. Ratings of functional ability of the affected upper limb of each participant were assessed prior to the study using the Motor Activity Log (MAL).2 Participation in the experiments did not require subjects to modify their own rehabilitation activities. Ethical approval for the study was obtained from the Auckland Ethics Committee. Protocol The rst neurophysiological and motor impairment assessments were completed one day prior to initiation of the training protocol (assessment

50

GN Lewis and WD Byblow

Details of the individual subjects Age 42 58 84 50 Sex M M M F Months since stroke 1 14 4 1 Lesion location left basal ganglia right anterior cerebral artery right parietal lobe right intracerebral haemorrhage right internal capsule right lenticular nucleus MALa amount 1.83 2.6 3.15 4.5 MALa how well? 1.67 2.57 2.5 3.86 Tasks B, S, P CI, PI, T B, S, P CI, S, P

Table 1 ID S1 S2 S3 S4

S5 S6
aMaximum

65 53

F M

47 30

1.61 0.4

1.61 0.3

B, S, P B, S, CI

score of 5. MAL, Motor Activity Log; M, male; F, female; B, block placement; S, simulated drinking; P, peg targeting; CI, cup inversion; PI, peg inversion; T, rapid transfer.

1). Over the next four-week period, the subjects completed a series of training sessions involving the performance of three speci c tasks. Five training sessions were completed per week, for a total of 20 training days. Each subject started out the training sessions using unimanual practice of the assigned tasks with the affected limb only, and then switched to bimanual training between days 7 and 13. To control for possible timing effects, a multiple baseline design was used to determine when each patient made the transition from unimanual to bimanual training: six protocols were designed starting on days 8 through 13, and each patient was randomly allocated to one of the unused protocols. Neurophysiological and motor impairment assessments of the upper limb were repeated at the end of the unimanual training period (assessment 2) and at the conclusion of the bimanual training period (assessment 3). During the training periods the subjects were seated in front of an adjustable table with an overhead rack (Figure 1). Each session involved repetitive practice of three tasks selected from: block placement, peg targeting, peg inversion, cup inversion, rapid transfer, and simulated drinking. The tasks were selected individually for each subject based on their functional ability, so that all patients completed tasks in which per-

formance de cits were noted in comparison to the unaffected side. The same three tasks were maintained for each patient over the duration of the study. At each training session 11 repetitions of each of the three tasks were performed. The rst trial of each task was designated the test trial and the remaining 10 were practice trials. The test trial was always performed unimanually with the affected upper limb and was video-taped from anterior and lateral aspects for later analysis. During the unimanual training sessions, the 10 practice trials were also performed with the affected upper limb only. At the bimanual training sessions the practice trials were performed with both upper limbs simultaneously. For example, in the block placement task the two upper limbs each picked up a block and placed them on two separate targets located on a box. Tasks Block placement: Subjects lifted a wooden block (7 5 5 cm) from the table and placed it on a target located on a 10-cm-high box (Figure 1). Subjects were instructed to lift and place the block in one movement whilst trying to be as accurate as possible. Peg targeting: Subjects picked up a peg (20 mm 5 mm length of dowel) from the table

Bilateral training intervention post stroke

51

adjustable overhead rack peg target

box block

Figure 1 The table used for performance of the tasks. The height of the overhead rack, position of the peg targets, and the location of the box could be adjusted for each individual.

and placed it in a target located on the underside of the overhead rack. The rack was aligned at eye height for each subject with the target set at shoulder width. Peg inversion: The peg was placed in the underside target of the rack at eye height. The subjects were instructed to pull the peg out and place it into the overhead side of the target, and then to return the peg to the starting position on the underside of the target. This resulted in the subject having to invert the peg as only one end tted in the target slots. Cup inversion: Subjects grasped the sides of an upside-down plastic cup from the table and placed it right-side-up on a target located on a 10-cm-high box, then returned the cup to the original location and position. Rapid transfer: Three small cylindrical containers (5 cm length, 2 cm diameter) were placed in a row (left-to-right) in a small box on the table. The containers were orientated horizontally and the box placed at a comfortable arm reach for the subject. A second box was placed on the table in line with the

rst but approximately 20 cm closer to the subject. The subjects task was to pick up each container individually and place it in a slot in the second box with the container in a vertical orientation. Once all three containers had been transferred the subject then placed each one back in its original position in the rst box. The transfer task was to be completed as fast as possible. Simulated drinking: A plastic cup was placed within arm reach on the table and the patients were instructed to pick up the cup and raise it to their mouth as though they were drinking from it. Neurophysiological assessments Using TMS, investigations were made of the excitability of the contralateral motor pathway from the affected hemisphere to the impaired limb and the ipsilateral pathways from the nonaffected hemisphere to the impaired limb. TMS was administered using a single Magstim 200 stimulator (Magstim Company, Dyfed, UK) and a gure-of-eight coil (outside diameter of each coil = 70 mm). The stimulating coil was orien-

52

GN Lewis and WD Byblow spot in the nonaffected hemisphere and the four sites immediately surrounding this. Eight stimuli were delivered over each of these ve locations. Electromyography EMG recordings were obtained from the two target muscles using 10-mm Ag/AgCl surface electrodes (Hydrospot, Physiometrix Inc., MA, USA). Following standard skin preparation, ECR electrodes were placed 2 cm apart on the belly of the muscle in the proximal forearm. For the FDI muscle, one electrode was placed over the metacarpo-phalangeal joint and the second over the belly of the muscle. Triggering of EMG collection was initiated 20 ms prior to stimulus onset. EMG signals were ampli ed using Grass P511AC ampli ers (Grass Instrument Division, RI, USA) and bandpass ltered (301000 Hz). Signals were sampled at 4000 Hz using a MacLab A/D acquisition system (ADInstruments, Castle Hill, NSW) and stored to disc for further analysis. Motor impairment ratings The motor impairment of the affected limb of all patients was assessed at each of the three test sessions using the hand and forearm sections of the Fugl-Meyer10 assessment scale. Data processing and analysis Analysis of task performance The video recordings of each of the test trials were placed in a random order and viewed by an independent rater who provided a measure of task performance for each trial. Motor performance of each task was assessed using six criteria: movement speed, movement smoothness, path directness, intralimb synchrony, quality of grasp, and joint angles at target.1,7 For the block placement and cup inversion tasks a measure of task accuracy was also given. For each variable analysed, the rater placed a mark at an appropriate position along a 10-cm line indicating their assessment of that variable, such that ratings of poor performance were indicated at the far lefthand end and perfect measures indicated at the right-hand end of the line. Perfect performance was de ned as equivalent to task performance by the nonaffected limb (one trial recorded at the end of the four-week training protocol). The tick

tated tangential to the scalp surface with the handle angled 45 from the midline and pointing backwards, such that the induced current ow was in a posterioranterior direction across the motor strip. Subjects wore a cotton cap with pre-marked grid locations, aligned with the centre of the cap over the subjects vertex. The hot spot for the extensor carpi radialis (ECR) and rst dorsal interosseous (FDI) muscles was located in the affected hemisphere by determining the scalp location eliciting motor evoked potentials (MEPs) of the largest amplitude in the two muscles. Resting excitability thresholds of the ECR and FDI muscles were determined as the lowest stimulus intensity that generated responses of greater than 50 V in at least four of a train of eight stimuli. Contralateral responses from stimulation over affected hemisphere Maps of cortical representation were created for the ECR and FDI muscles of the impaired limb while stimulating over the affected hemisphere. Stimulation intensity for mapping was set to 120% of the highest rest threshold of the two muscles. Stimuli were delivered at 21 grid locations surrounding the hot spot (5 5 grid without the corner sites). Six stimuli were delivered over each location. Ipsilateral responses from stimulation over intact hemisphere Ipsilateral responses to stimulation were recorded in the ECR and FDI muscles of the affected upper limb while the coil was held over the nonaffected hemisphere. Stimulus intensity was set to 150% of the highest rest threshold of the two muscles. If responses could not be elicited during contralateral stimulation in relaxed conditions (i.e., if rest threshold could not be determined), stimulus intensity was set to 75% of maximum stimulator output. During stimulation subjects maintained a 510% maximum voluntary contraction (MVC) in the ECR and FDI muscles of the affected limb. Visual feedback of electromyographic (EMG) activity from the two muscles was provided to assist subjects in maintaining this level of activation. Stimuli were delivered over the equivalent grid location as the hot

Bilateral training intervention post stroke marks on each scale were rounded to the nearest centimetre to give score from 0 to 10, and the total score for each trial determined. Thus, each trial received a performance rating out of 60 (70 for block placement and cup inversion tasks). Cortical representation maps of affected hemisphere For each evoked response collected the root mean square (rms) amplitude of EMG activity 20 ms prior to the stimulus was determined and responses were removed from further analysis if EMG silence was not maintained in this period. All remaining responses at each of the grid locations were averaged and the maximum peak-topeak amplitude of the averaged response determined. Measures of map area for each muscle were provided by summing the number of sites at which an averaged MEP amplitude of greater than 30 V was elicited. Ipsilateral responses from stimulation over intact hemisphere A 20-ms interval of EMG activity located poststimulus was analysed to determine the presence of an evoked potential. The start of the post-stimulus interval was de ned as 2 ms prior to the onset latency of contralateral MEPs from the intact hemisphere. Ipsilateral MEPs (iMEPs) were deemed present if the post-stimulus rms value was greater than 2.5 times the pre-stimulus value, and an obvious response was evident in visual inspection of the recording. Statistical analysis Analysis of trends in the ratings of task performance was carried out by visual inspection as well as statistically using time series analysis to detect signi cant shifts in data trends (FreeFore Professional, Automatic Forecasting Systems Inc.). Due to the small sample size, motor impairment ratings and the neurophysiological data were analysed using a nonparametric KruskalWallis test with the factor of session. Post-hoc investigations of signi cant effects were completed using a least signi cant difference between ranks test. The level of signi cance was set at 0.05. Results

53

During resting conditions, MEPs were only able to be elicited from the affected hemisphere of two of the six subjects at the initial test session. Responses from the affected musculature of a third subject (S4) were obtained from assessment 2 onwards. Two of the participants, S1 and S2, were only able to complete 18 and 19 of the 20 practice days, respectively. Task performance The ratings of task performance for each subject are presented in Figure 2. Most subjects demonstrated a gradual improvement in task performance ratings during the initial unimanual practice period. In a limited number of tasks an additional improvement in performance ratings was noted at the initiation of the bimanual training period. However, the bene cial effect of bilateral training was not consistent between tasks within a subject, and the performance of the majority of tasks was not signi cantly altered with the bilateral protocol. Statistically, the only signi cant level effects detected within two days of the switch to bimanual training were found in the simulated drinking task of S1, rapid transfer task of S2, and the peg targeting and simulated drinking tasks by S3 (all p < 0.01). Interestingly, the introduction of bilateral training appeared to reduce the progression in performance of the block placement task in subject S3. In con rmation of this, a signi cant level effect was found for this task (p < 0.05). Motor impairment ratings The Fugl-Meyer scores of all subjects are shown in Figure 3 for each of the three test sessions. There was no difference between the ratings across the three sessions (p = 0.5). Neurophysiological assessments Contralateral responses from stimulation over the affected hemisphere Due to the inability to elicit MEPs in the affected musculature of three subjects, maps of cortical representation were only completed for S1 and S2, and for the nal two assessments for S5 (see Table 2). Group analyses were not completed for these data.

54

GN Lewis and WD Byblow

Bilateral training intervention post stroke 55

Figure 2 Daily ratings of task performance for each subject. The acute (S1, S3, S4) and chronic (S2, S5, S6) subjects have been grouped separately. The dotted lines indicate the transition to bimanual practice sessions. Trendlines for the uni- and bimanual practice periods are shown by the solid lines. Signi cant level effects are indicated by the * symbol above the corresponding session (all p < 0.01).

56

GN Lewis and WD Byblow

Figure 3 Fugl-Meyer ratings for the individual subjects at the three test sessions. Ratings are out of a maximum of 32. The dashed line indicates the group average. Table 2 Rest threshold and map area results in the affected limb in response to stimulation over the affected hemisphere. Evoked potentials were obtained in three subjects only Subject Rest threshold (%) S1 S2 S4 53 43 Session 1 Map area FDI 17 16 Map area ECR 13 Rest threshold (%) 65 46 61 Session 2 Map area FDI 20 19 Map area ECR 1 14 Rest threshold (%) 75 38 72 Session 3 Map area FDI 18 21 12 Map area ECR 5 17

FDI, rst dorsal interosseus; ECR, extensor carpi radialis.

Surprisingly, S1 demonstrated a marked reduction in ECR map area from assessment 1 to assessment 2, despite improved ratings in functional ability of the wrist joint. Map area of the FDI muscle remained reasonably consistent across the four-week period. In subject S2, responses could only be elicited from the FDI muscle while at rest. Map area of the affected hemisphere increased moderately from assessment 1 to assessment 2, and again from assessment 2 to assessment 3. Evoked potentials could not be elicited from either target muscle in S4 at the initial assessment, however responses became evident in the ECR muscle after the unimanual training period and in the FDI muscle after the bimanual training period. A larger map area of the ECR muscle was evident in the nal session in this subject compared to the second assessment session.

Ipsilateral responses from stimulation over the intact hemisphere The number of ipsilateral responses elicited from both muscles combined is shown for each subject in Figure 4. Ipsilateral MEPs were seen in both muscles of all but one subject (S6) in all three test sessions. Overall, the total number of ipsilateral responses elicited was not signi cantly different between the three sessions (p = 0.8). Discussion This study examined the neurophysiological and behavioural adaptations of poststroke hemiparetics in response to a four-week bilateral training intervention. The results of the investigation provide limited support for the use of bilateral training therapy in the rehabilitation of

Bilateral training intervention post stroke

57

Figure 4 The total number of ipsilateral responses (iMEPs) evoked by stimulation over the intact hemisphere. Responses from both muscles have been combined. The dashed line indicates the group average.

hemiplegic individuals in that most of the tasks that were analysed did not demonstrate an additional bene cial effect of the bilateral practice sessions. The results of the neurophysiological studies investigating alterations in cortical excitability were largely inconclusive due to the limited number of individuals from whom responses could be evoked. The study had a number of limitations, including the small sample of subjects tested, the large differences in lesion location, functional ratings and time since stroke onset between subjects at the initiation of the study, and the dif culty of eliciting MEPs via stimulation over the affected cortical hemisphere. These, as well as the relatively low intensity and

duration of the training protocol, are all likely to contribute to the lack of consistent ndings. Behavioural adaptations to bilateral training While the bimanual intervention appeared to be bene cial for the performance of certain tasks in some subjects, the marked, immediate facilitatory effect reported by Mudie and Matyas1,7 was not evident. In most of the tasks performed by the well-recovered patients there was no real additional bene cial effect of bilateral practice. When a positive in uence of the bilateral intervention was suggested it tended to be in tasks with lower performance scores of the subjects with moderate motor de cits, similar to that reported in earlier studies.1 It may be of consequence that the peg targeting task, in which more reliable facilitatory effects were noted, has a large involvement of proximal musculature compared to the other tasks (i.e., the activation of proximal muscles is required to direct the movements of these tasks rather than simply acting as stabilizers). Given the larger contribution of bilaterally descending pathways to proximal (postural) musculature it may not be surprising that movements comprising activation of these particular muscles may pro t from bilateral training interventions. The results of the block placement task for S3 and, to a lesser extent, S1 were rather interesting. Subject S3 displayed an immediate and

Clinical messages A bilateral training protocol, involving simultaneous performance of upper limb tasks, elicited limited improvements in motor output in a group of poststroke hemiparetics. Bilateral training protocols may be more effective in enhancing the performance of simple tasks involving proximal muscles in more severely affected hemiparetic patients.

58

GN Lewis and WD Byblow time, although map area did increase in the FDI muscle of the same subject and in both the target muscles of S4. Other studies using TMS to examine motor responses in stroke patients during recovery have reported a similar increase in the area of representation of affected musculature with the recovery of motor function. 14 This is thought to re ect an increase in the excitability of neuronal tissue in the affected hemisphere as recovery progresses. Enlargements in the area of cortical representation have also been reported in individuals participating in constraint induced therapies (CIT) in which patients are forced to increase the use of their impaired limbs in everyday life.15 The descending motor pathways from the intact hemisphere may also play a role in the recovery of function after stroke. Often, a heightened excitability of the nonaffected hemisphere is reported poststroke, 14 and balancing of excitability levels between the two hemispheres is seen as good indicator for recovery of function.14,16 Ipsilateral responses in the affected musculature from stimulation over the intact hemisphere were infrequent in the current study but were obvious in all but one patient who was unable to voluntarily activate the two target muscles. While the group results were not signi cant, subject S1 did demonstrate an increased number of iMEPs at the nal assessment. It is dif cult to say whether or not the increased number of ipsilateral responses in S1 arose directly in response to the bilateral training protocol. This subject demonstrated a positive in uence of the bilateral paradigm in one the three tasks assigned, but either no effect or a negative in uence on performance was noted in the remaining two tasks. The neurophysiological ndings presented above do not appear to indicate a marked in uence of the bilateral training intervention on the descending motor pathways that were investigated. It is possible that the enhancements in task performance over the four-week period arose through the involvement of other descending motor tracts that were not investigated, such as the reticulospinal or rubrospinal pathways. The fact that performance improvements differed across the three tasks completed by each subject may indicate that behavioural adaptations did not arise through a general change in motor path-

marked decline in performance upon the introduction of bilateral training, while the steady improvement of S1 in the same task was reduced in response to the bilateral intervention. During the initial, unimanual period both of these subjects employed a dissimilar, yet still functional, grasp and movement pattern compared to that seen in the noninvolved arm. After the introduction of bilateral practice the two subjects attempted to mimic the movement pattern of the nonimpaired side during the test trials, however they were unable to do so successfully. The Fugl-Meyer assessments were completed to gauge any general alterations in upper limb function that may have occurred in response to the training periods. Subject S3 appeared to demonstrate a relative reduction in impairment ratings after the bilateral training period, and also noted improved performances after bilateral training in two of the three tasks completed. There were little remaining changes in the ratings evident in the other subjects. This is not surprising given the lack of marked performance changes in response to the training intervention. Neurophysiological adaptations to bilateral training Given that there were no marked or consistent changes in the behavioural measures of the patients in response to the bilateral intervention, it may be expected that there would also be a lack of marked alterations in the activity or excitation levels of the motor pathways analysed. The neurophysiological measures may, however, provide some insight into any neural adaptations in response to bilateral training that are not manifest in functional performance measures. An increase in motor threshold is one of the most common ndings after stroke,1113 however Turton and colleagues13 reported the presence of MEPs in distal upper limb musculature of all subjects who were capable of voluntary muscle activation. The lack of responses in the subjects in the current study is therefore surprising given that all but one subject (S6) were able to voluntarily activate the target muscles of the affected limb, although it may be explained by the differences in the coil utilized. In the ECR muscle of S1 there was a reduction in the number of map locations eliciting discernible responses over

Bilateral training intervention post stroke way excitability. Rather, performance enhancements may be related to superior intralimb coordination with practice, or the adoption of more functional strategies or muscle synergies to complete the assigned tasks. Acknowledgements The authors are extremely grateful for the assistance of Dr Samir Anwar, Denise Lyness, Carolann Murray and Louise Jackman (Rehab Plus, Auckland Healthcare) and Yvette Baker (Stroke Unit, Auckland Hospital) in the recruitment of patients for this study. We would also like to thank Jim Stinear and Cathy Stinear for their assistance in the laboratory. GN Lewis is supported by a scholarship from the Foundation for Research, Science and Technology. This project was funded in part by the University of Auckland Graduate Research Fund and a University of Auckland Staff Research Grant. References
1 Mudie MH, Matyas TA. Upper extremity retraining following stroke: effects of bilateral practice. J Neurol Rehabil 1996; 10: 16784. 2 Taub E, Miller NE, Novack TA et al. Technique to improve chronic motor de cit after stroke. Arch Phys Med Rehabil 1993; 74: 34754. 3 Cunningham CL, Phillips Stoykov ME, Walter CB. Bilateral facilitation of motor control in chronic hemiplegia. Acta Psychol 2002; 110 : 32137. 4 Volman MJM, Wijnroks A, Vermeer A. Bimanual circle drawing in children with spastic hemiparesis: effect of coupling modes on the performance of the impaired and unimpaired arms. Acta Psychol 2002; 110 : 33956. 5 Steenbergen B, Hulstijn W, de Vries A, Berger M. Bimanual movement coordination in spastic hemiparesis. Exp Brain Res 1996; 110 : 9198. 6 Steenbergen B, van Thiel E, Hulstijn W, 13

59

9 10

11

12

14

15 16

Meulenbroek RGJ. The coordination of reaching and grasping in spastic hemiparesis. Hum Mov Sci 2000; 19: 75105. Mudie MH, Matyas TA. Can simultaneous bilateral movement involve the undamaged hemisphere in reconstruction of neural networks damaged by stroke? Disabil Rehabil 2000; 22: 2337. Whitall J, McCombe Waller S, Silver KHC, Macko RF. Repetitive bilateral arm training with rhythmic auditory cueing improves motor function in chronic hemiparetic stroke. Stroke 2000; 31: 239095. Cauraugh JH, Kim S. Electromyogram-triggered neuromuscular stimulation and bilateral movements. Stroke 2002; 33: 158994. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. I. A method for evaluation of physical performance. Scand J Rehabil Med 1975; 7: 1331. Byrnes ML, Thickbroom GW, Phillips BA, Wilson SA, Mastaglia FL. Physiological studies of the corticomotor projection to the hand after subcortical stroke. Clin Neurophys 1999; 110: 48798. Catano A, Houa M, Noel P. Magnetic transcranial stimulation: clinical interest of the silent period in acute and chronic stages of stroke. Electroencephalogr Clin Neurophysiol 1997; 105 : 29096. Turton A, Wroe S, Trepte N, Fraser C, Lemon RN. Contralateral and ipsilateral EMG responses to transcranial magnetic stimulation during recovery of arm and hand function after stroke. Electroencephalogr Clin Neurophysiol 1996; 101 : 31628. Cicinelli P, Traversa R, Rossini PM. Post-stroke reorganisation of brain motor output to the hand: a 24 month follow-up with focal magnetic transcranial stimulation. Electroencephalogr Clin Neurophysiol 1997; 105 : 43850. Liepert J, Bauder H, Miltner WHR, Taub E, Weiller C. Treatment-induced cortical reorganisation after stroke in humans. Stroke 2000; 31: 121016. Traversa R, Cicinelli P, Filippi M et al. A method to monitor motor cortical excitability in human stroke through motor evoked potentials. Brain Res Protoc 1999; 4: 4448.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.