https://doi.org/10.1140/epjs/s11734-025-01555-6
Regular Article
Stretchable high-density surface electromyography electrode patch assisted with machine learning for silent speech recognition
1
Department of Rehabilitation Medicine, The First Affiliated Hospital School of Medicine, Zhejiang University, 310003, Hangzhou, China
2
Key Laboratory of Soft Machines and Smart Devices of Zhejiang Province, State Key Laboratory of Brain-Machine Intelligence, Department of Engineering Mechanics, Zhejiang University, 310027, Hangzhou, China
3
Huanjiang Laboratory, 311899, Zhuji, China
Received:
17
November
2024
Accepted:
27
February
2025
Published online:
14
March
2025
Approximately 400 million people worldwide suffer from speech disorders, which pose significant challenges to their daily communications. Therefore, it is of great significance to develop a wearable speech recognition system to achieve assistive communication and human–computer interaction for people with speech impairments. Here, we report a stretchable high-density surface electromyography (HD-sEMG) electrode patch designed to capture HD-sEMG signals for silent speech recognition. The patch, features 32 working electrodes on a soft polyurethane substrate, which can conformably cover most of the throat area and maintain stable contact with the skin even during involuntary head movements. Characterization results show that the electrode has excellent adhesion, low interfacial impedance, and strong resistance to motion artifacts, enabling high-fidelity signal acquisition. Using these signals, a neural network classification model with a multi-layer convolutional architecture achieved an overall classification accuracy of 96.09% for common words and short sentences, indicating robust feature extraction with minimal misclassification. Visualization of sEMG signal clusters using t-distributed stochastic neighbor embedding (t-SNE) further validated the model’s ability to distinguish types of unvoiced speech, highlighting its generalizability and precision. This study demonstrates that the stretchable HD-sEMG electrode patch serves as a promising wearable interface for silent speech recognition and has great potential in assistive communication for speech-impaired users.
Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to EDP Sciences, Springer-Verlag GmbH Germany, part of Springer Nature 2025
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
