Objective: The compound muscle action potential (CMAP) scan is a useful technique for examination of neuromuscular disorders. The objective of this study is to develop a novel analysis of CMAP scans from the perspective of information theory. Methods: A novel index parameter called CMAP distribution index (CDIX) was developed to characterize CMAP scan based on calculation of the information entropy. The performance of CDIX was evaluated using CMAP scan data from healthy control and spinal cord injury (SCI) subjects, and compared with D50 and MScanFit motor unit number estimation (MUNE). Results: CDIX was significantly lower for the SCI subjects compared with the healthy control subjects (p < 0.001). A significant correlation ( ${\mathbf{R}}^2$ = 0.58, p < 0.001) was found between CDIX and MScanFit MUNE. Among all tested parameters (maximum CMAP, D50, MScanFit MUNE and CDIX), CDIX achieved the smallest relative width of the overlapping zone (WOZ%) between SCI and healthy control subjects. Conclusion: CDIX can be inferred as a useful index reflecting motor unit loss and muscle fiber reinnervation changes.

Deaf and hearing-impaired people always face communication barriers. Non-invasive surface electromyography (sEMG) sensor-based sign language recognition (SLR) technology can help them to better integrate into social life. Since the traditional tandem convolutional neural network (CNN) structure used in most CNN-based studies inadequately captures the features of the input data, we propose a novel inception architecture with a residual module and dilated convolution (IRDC-net) to enlarge the receptive fields and enrich the feature maps, applying it to SLR tasks for the first time. This work first transformed the time domain signal into a time–frequency domain using discrete Fourier transformation. Second, an IRDC-net was constructed to recognize ten Chinese sign language signs. Third, the tandem CNN networks VGG-net and ResNet-18 were compared with our proposed parallel structure network, IRDC-net. Finally, the public dataset Ninapro DB1 was utilized to verify the generalization performance of the IRDC-net. The results showed that after transforming the time domain sEMG signal into the time–frequency domain, the classification accuracy (acc) increased from 84.29% to 91.70% when using the IRDC-net on our sign language dataset. Furthermore, for the time–frequency information of the public dataset Ninapro DB1, the classification accuracy reached 89.82%; this value is higher than that achieved in other recent studies. As such, our findings contribute to research into SLR tasks and to improving deaf and hearing-impaired people’s daily lives.

A surface electromyography (EMG) analysis was performed in this study to examine central neural and peripheral muscle changes after a spinal cord injury (SCI). A linear electrode array was used to record surface EMG signals from the biceps brachii (BB) in 15 SCI subjects and 14 matched healthy control subjects as they performed elbow flexor isometric contractions from 10% to 80% maximum voluntary contraction. Muscle fiber conduction velocity (MFCV) and BB EMG–force relation were examined. MFCV was found to be significantly slower in the SCI group than the control group, evident at all force levels. The BB EMG–force relation was well fit by quadratic functions in both groups. All healthy control EMG–force relations were best fit with positive quadratic coefficients. In contrast, the EMG–force relation in eight SCI subjects was best fit with negative quadratic coefficients, suggesting impaired EMG modulation at high forces. The alterations in MFCV and EMG–force relation after SCI suggest complex neuromuscular changes after SCI, including alterations in central neural drive and muscle properties.

Weakness is a common symptom after a stroke. This study aims to depict weakness distribution among forearm muscles given that joints in the upper limb are usually driven by a group of muscles. Multi-channel electromyography (EMG) was applied to assess the muscle group, and an EMG-based index was proposed to quantify the weakness of individual muscles. By using this method, four weakness distribution patterns were observed in extensor muscles from five out of eight subjects after stroke. Complex weakness distribution patterns were observed in flexor muscles from seven out of the eight subjects when they performed grasp, tripod pinch, and hook grip. The findings can help determine the weak muscles in a clinic and facilitate the development of appropriate interventions in stroke rehabilitation targeting specific muscle weakness.

Recently, robot-assisted rehabilitation has emerged as a promising solution to increase the training intensity of stroke patients while reducing workload on therapists, whilst surface electromyography (sEMG) is expected to serve as a viable control source. In this paper, we delve into the potential of deep learning (DL) for post-stroke hand gesture recognition by collecting the sEMG signals of eight chronic stroke subjects, focusing on three primary aspects: feature domains of sEMG (time, frequency, and wavelet), data structures (one or two-dimensional images), and neural network architectures (CNN, CNN-LSTM, and CNN-LSTM-Attention). A total of 18 DL models were comprehensively evaluated in both intra-subject testing and inter-subject transfer learning tasks, with two post-processing algorithms (Model Voting and Bayesian Fusion) analysed subsequently. Experiment results infer that for intra-subject testing, the average accuracy of CNN-LSTM using two-dimensional frequency features is the highest, reaching 72.95%. For inter-subject transfer learning, the average accuracy of CNN-LSTM-Attention using one-dimensional frequency features is the highest, reaching 68.38%. Through these two experiments, it was found that frequency features had significant advantages over other features in gesture recognition after stroke. Moreover, the post-processing algorithm can further improve the recognition accuracy, and the recognition effect can be increased by 2.03% through the model voting algorithm.

This study presents a novel compound muscle action potential (CMAP) examination of motor unit changes in paretic muscle post stroke. CMAP scan of the first dorsal interosseous (FDI) muscle was performed bilaterally in 16 chronic stroke subjects. Various parameters were derived from the CMAP scan to examine paretic muscle changes, including CMAP amplitude, D50, step index (STEPIX) and amplitude index (AMPIX). A significant decrease in CMAP amplitude and STEPIX was observed in paretic muscles compared with contralateral muscles (CMAP amplitude: paretic (9.0±0.5) mV, contralateral (11.3±0.9) mV, P=0.024; STEPIX: paretic 101.2±7.6, contralateral 121.9±6.5, P=0.020). No significant difference in D50 and AMPIX was observed between the paretic and contralateral sides (P>0.05). The findings revealed complex paretic muscle changes including motor unit degeneration, muscle fiber denervation, reinnervation and atrophy, providing useful insights to help understand neuromuscular mechanisms associated with weakness and other functional deterioration post stroke. The CMAP scan experimental protocols and the applied processing methods are noninvasive, convenient, and automated, offering practical benefits for clinical application.