Abstract:
Surface electromyography (sEMG) based neuro-interface technology has been widely researched in stroke rehabilitation. However, sEMG features of stroke patients are pathol...Show MoreMetadata
Abstract:
Surface electromyography (sEMG) based neuro-interface technology has been widely researched in stroke rehabilitation. However, sEMG features of stroke patients are pathologically changed, which may greatly decrease sEMG decoding performance. Overall estimation of sEMG changes in stroke patients would benefit the clinical usage of sEMG technology in stroke rehabilitation. In this study, we collected sEMG signals from both stroke patients (n=11) and able-bodied individuals (n=11). The sEMG patterns and movement decoding performance were analyzed across different subjects. Relationships between sEMG decoding accuracy and Upper-FMA (uFMA) scores in stroke patients were evaluated. We find that there is a significant correlation between sEMG decoding accuracy and uFMA scores (Spearman r=0.75, p<0.01), and then we separate stroke patients into Low-FMA group and High-FMA group with a threshold of uFMA=22. The High-FMA group share more similar activation in MAV maps with Healthy group, while Low-FMA group shows different compensatory peaks on the upper arms. Our study finds that sEMG signals could be generally used for movement decoding in stroke patients with higher uFMA scores (>22). High-FMA group share similarity in muscle activation with Health group. This finding provides a possible way for better adaption of sEMG signals in stroke rehabilitation and for selection of suitable patients in sEMG-based active rehabilitation.
Published in: 2021 27th International Conference on Mechatronics and Machine Vision in Practice (M2VIP)
Date of Conference: 26-28 November 2021
Date Added to IEEE Xplore: 07 January 2022
ISBN Information: