28 June 2023 in IEEE Transactions on Neural Systems and Rehabilitation Engineering
This study aimed to evaluate the utility of sample entropy (SampEn) of surface electromyographic signals (sEMG) in quantifying neuromuscular changes after spinal cord injury (SCI). Using a linear electrode array, sEMG signals were obtained from the biceps brachii muscles of 13 healthy control subjects and 13 SCI subjects during isometric elbow flexion muscle contraction at various constant force levels. The SampEn analysis was performed on both the representative channel (emerging the largest signal amplitude) and the channel over muscle innervation zone (IZ) determined by the linear array. The SampEn values were further averaged across muscle force levels to examine difference between SCI survivors and control subjects. Results showed that SampEn values after SCI had a significantly greater range than those of the control subjects at the group level. Abnormal increases or decreases in SampEn were also found after SCI at the individual subject level. In addition, a significant difference was found between the representative channel and the IZ channel. SampEn is a valuable indicator for identifying neuromuscular changes after SCI, and the effect of the IZ on the sEMG examination is particularly noteworthy. The approach presented in this study may facilitate the development of appropriate rehabilitation methods to enhance motor recovery.
6 March 2023 in IEEE Transactions on Neural Systems and Rehabilitation Engineering
Objective : This study aims to develop a flexible myoelectric pattern recognition (MPR) method based on one-shot learning, which enables convenient switching across different usage scenarios, thereby reducing the re-training burden. Methods : First, a one-shot learning model based on a Siamese neural network was constructed to assess the similarity for any given sample pair. In a new scenario involving a new set of gestural categories and/or a new user, just one sample of each category was required to constitute a support set. This enabled the quick deployment of the classifier suitable for the new scenario, which decided for any unknown query sample by selecting the category whose sample in the support set was quantified to be the most like the query sample. The effectiveness of the proposed method was evaluated by experiments conducting MPR across diverse scenarios. Results: The proposed method achieved high recognition accuracy of over 89% under the cross-scenario conditions, and it significantly outperformed other common one-shot learning methods and conventional MPR methods ( p < 0.01). Conclusion : This study demonstrates the feasibility of applying one-shot learning to rapidly deploy myoelectric pattern classifiers in response to scenario change. It provides a valuable way of improving the flexibility of myoelectric interfaces toward intelligent gestural control with extensive applications in medical, industrial, and consumer electronics.
27 April 2024 in Journal of Electromyography and Kinesiology