Deep Muscle EMG construction using A Physics-Integrated Deep Learning approach

2025-03-07

AI Summary

This study addresses the long-standing challenge of noninvasively acquiring electromyographic (EMG) signals from deep muscles. We propose a physics-informed hybrid neuromusculoskeletal model (NMM) that synergistically integrates biomechanical modeling with data-driven deep learning. The model takes surface EMG and kinematic data as inputs and explicitly encodes subject-specific musculoskeletal constraints—such as muscle architecture, moment arms, and joint dynamics—into its neural architecture, thereby overcoming limitations of purely data-driven or purely physics-based approaches. Built upon the physics-informed neural networks (PINN) framework, the method is rigorously validated using multi-source evidence, including OpenSim simulations and experimental measurements across five human subjects. Results demonstrate high-fidelity reconstruction of deep-muscle activation: reconstruction error is significantly reduced compared to conventional muscle synergy extrapolation (MSE); predicted joint torques closely match OpenSim simulations (RMSE < 0.12 N·m). This work delivers a generalizable, interpretable, and noninvasive solution for EMG-driven neuromusculoskeletal modeling.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Deep Neural Architectures and Foundation Models
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Electromyography (EMG)--based computational musculoskeletal modeling is a non-invasive method for studying musculotendon function, human movement, and neuromuscular control, providing estimates of internal variables like muscle forces and joint torques. However, EMG signals from deeper muscles are often challenging to measure by placing the surface EMG electrodes and unfeasible to measure directly using invasive methods. The restriction to the access of EMG data from deeper muscles poses a considerable obstacle to the broad adoption of EMG-driven modeling techniques. A strategic alternative is to use an estimation algorithm to approximate the missing EMG signals from deeper muscle. A similar strategy is used in physics-informed deep learning, where the features of physical systems are learned without labeled data. In this work, we propose a hybrid deep learning algorithm, namely the neural musculoskeletal model (NMM), that integrates physics-informed and data-driven deep learning to approximate the EMG signals from the deeper muscles. While data-driven modeling is used to predict the missing EMG signals, physics-based modeling engraves the subject-specific information into the predictions. Experimental verifications on five test subjects are carried out to investigate the performance of the proposed hybrid framework. The proposed NMM is validated against the joint torque computed from 'OpenSim' software. The predicted deep EMG signals are also compared against the state-of-the-art muscle synergy extrapolation (MSE) approach, where the proposed NMM completely outperforms the existing MSE framework by a significant margin.

Problem

• Estimating deep muscle EMG signals non-invasively
• Overcoming challenges in measuring deep muscle EMG
• Integrating physics-informed and data-driven deep learning

Innovation

• Physics-integrated deep learning for EMG estimation
• Neural musculoskeletal model (NMM) combines physics and data
• NMM outperforms muscle synergy extrapolation (MSE) approach