sEMG-Driven Physics-Informed Gated Recurrent Networks for Modeling Upper Limb Multi-Joint Movement Dynamics

2024-08-29

arXiv.org

AI Summary

Real-time, high-accuracy prediction of upper-limb multi-joint kinematic and dynamic variables—including joint angles, velocities, accelerations, external loads, and torques—remains challenging in exoskeletons and rehabilitation systems. Method: This paper proposes a Physics-informed Gated Recurrent Network (PiGRN), the first to embed biomechanical dynamic equations (i.e., torque–motion relationships) directly into a differentiable Gated Recurrent Unit (GRU), yielding an end-to-end trainable, physics-constrained neural architecture. PiGRN jointly extracts spatiotemporal features from surface electromyography (sEMG) signals and performs multi-task regression to predict full-order dynamical parameters. Results: Evaluated on elbow flexion–extension tasks across five subjects, PiGRN achieves RMSEs of 4.02%–11.40% and correlation coefficients of 0.87–0.98 for joint torque prediction across ten unseen movements—demonstrating substantial improvements in prediction accuracy, responsiveness, and cross-movement generalization for human–machine interfaces.

Technology Category

• Intelligent Robots: State Estimation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Learning with Manifolds

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Exoskeletons and rehabilitation systems have the potential to improve human strength and recovery by using adaptive human-machine interfaces. Achieving precise and responsive control in these systems depends on accurately estimating joint movement dynamics, such as joint angle, velocity, acceleration, external mass, and torque. While machine learning (ML) approaches have been employed to predict joint kinematics from surface electromyography (sEMG) data, traditional ML models often struggle to generalize across dynamic movements. In contrast, physics-informed neural networks integrate biomechanical principles, but their effectiveness in predicting full movement dynamics has not been thoroughly explored. To address this, we introduce the Physics-informed Gated Recurrent Network (PiGRN), a novel model designed to predict multi-joint movement dynamics from sEMG data. PiGRN uses a Gated Recurrent Unit (GRU) to process time-series sEMG inputs, estimate multi-joint kinematics and external loads, and predict joint torque while incorporating physics-based constraints during training. Experimental validation, using sEMG data from five participants performing elbow flexion-extension tasks with 0 kg, 2 kg, and 4 kg loads, showed that PiGRN accurately predicted joint torques for 10 novel movements. RMSE values ranged from 4.02% to 11.40%, with correlation coefficients between 0.87 and 0.98. These results underscore PiGRN's potential for real-time applications in exoskeletons and rehabilitation. Future work will focus on expanding datasets, improving musculoskeletal models, and investigating unsupervised learning approaches.

Problem

• Accurate estimation of joint movement dynamics
• Integration of biomechanical principles in neural networks
• Prediction of multi-joint kinematics from sEMG data

Innovation

• Physics-informed Gated Recurrent Network
• sEMG data processing
• Multi-joint dynamics prediction