To obtain access to full text of journal and articles you must register!
- Article name
- Neural network analysis of multichannel EMG signals: architectures, algorithms, and applications in the management of prosthetic limbs
- Authors
- Garafutdinov A. A., , time-v@yandex.ru, FSBEI HE "Kazan National Research Technical University named after A. N. Tupolev - KAI", Kazan, Russia
- Keywords
- electromyography / myointerface / prosthetics / machine learning / deep learning / CNN / LSTM / transformers / gesture recognition
- Year
- 2025 Issue 2 Pages 34 - 41
- Code EDN
- VNPKEQ
- Code DOI
- 10.52190/2073-2600_2025_2_34
- Abstract
- The article discusses modern methods of analyzing electromyographic (EMG) signals for controlling prosthetic limbs. Special attention is paid to the use of machine learning algorithms, including convolutional (CNN) and recurrent neural networks (LSTM), as well as transformers, to automatically extract spatiotemporal patterns in multichannel EMG data. A comparative analysis of classical and modern approaches is carried out, their advantages and limitations are discussed. The key stages of signal processing are considered: filtering, segmentation, feature extraction and classification. Examples of using open databases (NinaPro, BioPatRec) for training and validating models are given. It is shown that deep learning makes it possible to increase the accuracy of calculations.
- Text
- BUY for read the full text of article
- Buy
- 500.00 rub
