Convolutional Neural Networks for Time Series Data Processing Applicable to sEMG Controlled Hand Prosthesis




Convolutional Neural Network, Feature Engineering, Residual Neural Network, Spatial Feature, Surface Electromyography, Time Series Data


Surface electromyography (sEMG) signals are often used to control prosthetics, but accurately interpreting these stochastic signals remains challenging. Deep learning tools like convolutional neural networks (CNNs) have shown promise for complex classification problems, yet CNN applications for time series data are limited. This work explores adapting CNNs to sEMG time series for improved classification, addressing two questions: 1) Can a CNN trained on cross-subject data generalize without individualization? 2) Can a small individualized dataset sufficiently train an accurate control model? To investigate, sEMG data is formatted into images using handcrafted features, with pixels representing multichannel time series. A ResNet50 architecture is trained on two datasets: individual and cross-subject. Results show cross-subject models fail to provide accurate subject-specific control due to high inter-subject variability of sEMG. However, ResNet50 trained on individual data produces highly accurate offline and near real-time classification. The proposed method is also tested on an external dataset and compared to similar published methods, demonstrating strong performance. In summary, CNNs show promise for prosthetic control from sEMG but require individualized training data. The proposed data formatting and ResNet50 architecture can enable precise control from minimal data, overcoming barriers to clinical implementation. Further research into cross-subject generalizability is warranted to understand the sources of variability and improve model robustness.

Author Biographies

Golam Gause Jaman, Measurement and Control Engineering Research Center, Idaho State University

Golam Gause Jaman received the B.S. in electronics and telecommunication engineering from North South University, Dhaka, Bangladesh in 2013 and a B.S. in electrical engineering from Idaho State University, Pocatello, ID, USA in 2015. He is currently pursuing Ph.D. in engineering and applied science with measurement and control engineering emphasis at Idaho State University, Pocatello, ID, USA.

Since 2016, he is working as a Research Assistant with the department of mechanical engineering, Idaho State University, Pocatello, ID, USA. His research interest includes the development of upper limb motion classification using surface electromyography signals and machine learning. He is also currently engaged in reinforcement learning controller design research applicable in the field of additive and advanced manufacturing.

Marco Schoen, Measurement and Control Engineering Research Center, Idaho State University

Marco P. Schoen holds a B.S. degree in Mechanical Engineering from the Swiss College of Engineering at Muttenz, Switzerland, a M.S. in Mechanical Engineering from Widener University, Pennsylvania, USA, and a Ph.D. in Engineering Mechanics from Old Dominion University, Virginia USA.

He is a professor of the Department of Mechanical Engineering and currently serving as the director of the Measurement and Controls Engineering Research Center at Idaho State University. His research interests are in modeling and simulation, estimation theory, intelligent controls, adaptive controls, system identification, and machine learning with applications found in biomedical engineering, aerospace engineering, energy systems and advanced manufacturing.




How to Cite

Jaman, G. G. and Schoen, M. (2024) “Convolutional Neural Networks for Time Series Data Processing Applicable to sEMG Controlled Hand Prosthesis”, Technische Mechanik - European Journal of Engineering Mechanics, 44(1), pp. 47–60. doi: 10.24352/UB.OVGU-2024-053.