Effect of Botulinum Toxin Type A on EMG-based Motion Detection: a Pilot Study On Post-Stroke Patients

The outcome of robotic-assisted training can be enhanced by controlling the robot using the user's own motion. Electromyography (EMG) signals have been used to detect motions based on muscle activities. However, spasticity-induced involuntary muscle activities can be a main factor compromising the accuracy of motion detection for post-stroke patients. This study therefore aimed to quantify the effect of Botulinum toxin type A (BoNT-A) treatment on the accuracy of EMG-based motion detection for post-stroke patients with spasticity. Four stroke patients who received BoNT-A treatment on their flexor digitorum superficialis muscle due to moderate or severe spasticity were recruited in this study. High-density surface EMG signals of hand closing were recorded before BoNT-A treatment and two weeks afterwards. Threshold-based method was used to detect motions, and motion detection accuracy was calculated for each subject before and after BoNT-A treatment. Increased accuracy of motion detection was observed after treatment. The findings of this study provide evidence that muscle spasticity reduction by BoNT-A treatment may improve motion detection, and can improve the interaction experience of robot-assisted stroke rehabilitation by reducing the chance of unintended triggering.