E2CNN: An Efficient Concatenated CNN for Classification of Surface EMG Extracted from Upper-Limb

Surface electromyography are bioelectrical indicators that emerge during muscle contraction and have been widely used in a variety of clinical applications. Several prosthetic control applications can benefit from analysis based on the classification of sEMG signals. However, for the real-time application of upper-limb prosthesis, EMG-based systems need robust performance and rapid response behaviour. In this study, we propose an efficient concatenated convolutional neural network (E2CNN) for classification of sEMG extracted from upper-limb. We have tested and validated the performance of the proposed E2CNN on two datasets: a longitudinal dataset comprised of ten able-bodied (healthy) subjects as well as six transradial amputee subjects and spanned the data collected for a period of seven days; and the publicly available NinaPro DB1 dataset. The raw sEMG signals are converted to Log-Mel (LM) spectrograms. This model combines input layers with the output of each convolutional block using concatenation layers. The proposed efficient concatenated CNN (E2CNN) when applied to log-Mel spectrogram-based images, provides a good response time with high-performance accuracy of 98.31% ± 0.5% and 97.97% ± 1.41% for both able-bodied and amputee subjects. When applied to NinaPro DB1, the proposed E2CNN has attained a mean accuracy of 91.27%, an increase by 24.67% with respect to baseline CNN model. The results show that the achieved results are comparable to those obtained using SSAE and other CNN models; however, E2CNN is associated with reduced training and prediction time, making it a potential candidate for real-time classification of sEMG based on LM spectrogram images.