An Attention-based Bidirectional LSTM Model for Continuous Cross-subject Estimation of Knee Joint Angle during Running from sEMG Signals

Running is an essential locomotion activity that plays a critical role in everyday life and exercise activities and may be impeded by joint disease and neurological impairments. Accurate and robust estimation of joint kinematics via surface electromyogram (sEMG) signals provides a human-machine interaction-based method that can be used to adequately control rehabilitation robots while performing complex movements such as running for motor function restoration in affected persons. To this end, this paper proposes a novel deep learning-based model (AM-BiLSTM) that integrates an attention mechanism (AM) and a bidirectional long short-term memory (BiLSTM) network. The proposed method was evaluated using knee joint kinematic and sEMG signals of fourteen subjects who performed running at 2 m/s speed. The proposed model’s generalizability was tested for within- and cross-subject scenarios and compared with standard LSTM and multi-layer perceptron (MLP) networks in terms of normalized root-mean-square error and correlation coefficient evaluation metrics. Based on the statistical tests, the proposed AM-BiLSTM model significantly outperformed the LSTM and MLP methods in both within- and cross-subject scenarios (p<0.05) and achieved state-of-the-art performance.

© 20XX IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.