The directionality of interjoint neural coupling in the impaired arm post stroke

To date, research has primarily focused on identifying and analyzing abnormal coupling in the upper extremity following stroke during gross movements such as reaching. Yet, little is known about the direction of abnormal coupling in conjunction with its degree between joints—muscles—of interest. In this study, we address the directionality of pathological neural couplings between joints, and variability in the degree of them, in the upper extremity post stroke. To evaluate the direction of interjoint coupling, joint motion and muscle activity in an involuntary mode while another joint—muscle—was voluntarily activated were assessed in comparison to those made in a volitional mode, focusing on the shoulder, elbow and wrist joints. Overall, involuntary activation of the wrist flexor in the stroke group is observed when muscles of a more proximal joint voluntarily activate (p < 0.1 and p <0.05), while activation of the wrist flexor does not lead to involuntary activation of the shoulder muscles and elbow flexor, in comparison to the healthy group. In particular, in the stroke group who is assumed to have a severe loss of the corticospinal system, the wrist extensor is subordinately activated by those about the other joints voluntarily activated (p < 0.05) while it does not cause activations of those about the other joints. These findings indicate that stroke-caused synergies could be directional, implying the existence of a cortical source and/or neural pathway for a muscle that is not paired with that for other muscles.


28
Stroke can be characterized by impaired motor coordination as it evolves to the stage following  distal muscles are bundled together (Fetz and Cheney, 1980). Due probably to these facts, we have 45 observed inconclusive results of stroke regarding the severity gradient of motor deficits across the 46 impaired arm(Colebatch and Gandevia, 1989; Beebe and Lang, 2008).

47
Though it is obvious that neural couplings between the joints of the upper extremity (through their that in a pair of muscles, one muscle is activated in an involuntarily manner when the other muscle is activated, but the other muscle is not activated when the muscle is activated. These phenomena 58 would imply the existence of a cortical source and/or neural pathway for the joint or muscle that 59 is not paired with that for the other joint or muscle. It is worth investigating the direction of 60 coupling as well as the severity of coupling between joints of a stroke-affected arm to gain an 61 insight into diffuse projection of cortical sources and routes that convey neural drive to a certain 62 muscle or joint.

63
In this study, we investigate the direction of coupling as well as the extent of coupling between 64 joints and between muscles of the upper limb of a population with moderate-to-severe impairment

71
We strive to distinguish the features of stroke-caused coupling based on the severity of     The initial configuration of the arm was at shoulder adduction of 45°, shoulder flexion of 45°,

112
The experiment consisted of three sessions. Participants were asked to voluntarily move only one 113 joint (the target joint) with the other joints (non-target joints) resting, as depicted in Fig. 1. The

114
shoulder, elbow and wrist joints of the exoskeleton were able to be moved freely with minimized mechanical impedance. The target joint was selected in the order of the shoulder, elbow and wrist 116 joints. The movement and its speed at each joint were designed as 7~9 cycles at self-selected 117 comfortable speeds. Break times were inserted to mitigate the fatigue effect during the experiment.

118
Data processing

119
The angle and EMG data were time synchronized using a pulse signal sent to different personal 120 computers used during the data acquisition. Raw angle data were used themselves without filtering.

121
Raw EMG data were high-pass filtered at 20 Hz, de-meaned, rectified, and low-pass filtered at 5 We employed multivariate analysis of variance (MANOVA) with Bonferroni corrections to 147 investigate differences between groups in ratios for motion coupling and muscle coupling. For 148 motion coupling, two-way MANOVA was performed on the six ratios, with group as a between-149 subject variable and with direction as a within-subject variable. For muscle coupling, one-way 150 MANOVA was performed on the 24 ratios with group as a between-subject variable. We used 151 one-tailed t-tests to determine whether ratios were significantly lower than 1. The data that support the findings of this study are available from the corresponding author on 160 reasonable request.

163
Mainly, we found that the wrist flexor tends to involuntarily activate (marginally) significantly 164 more in the stroke group when the shoulder or elbow voluntarily moves, in comparison to the 165 healthy group. Meanwhile, it is interesting that activation of the wrist flexor does not lead to 166 involuntary activation of AD, PD and Bi, which would imply that abnormal coupling is directional.

167
In the stroke group who is assumed to be with a severe loss of the corticospinal system, the wrist  Results on the whole stroke participants observe no effects of movement direction (p > 0.9) and interaction (p > 0.9), indicating no 174 significant differences in the extent of motion coupling between inward movements and outward 175 movements (refer to Fig. 3(a)). However pairwise comparisons between groups report that the

187
MANOVA reports no main effect of group in ratios for muscle activation coupling (p > 0.6).

188
Meanwhile, the stroke group shows significant greater ratios for the PD muscle for the cases of

196
One-tailed t-tests revealed that ratios for all cases others than three cases are significantly lower

276
The phenomenon that we observe ratios not significantly less than 1 in the biceps and triceps 277 across healthy participants is due partially to an effort of stabilization of the forearm through an  voluntarily moves is not significantly less than that of its voluntary motion (refer to Fig. 5).

304
The results of muscle activation coupling in the stroke group A show a greater extent of abnormal

314
The directional neural couplings we observed at least in this study probably originate from the