Classifying muscle parameters with artificial neural networks and simulated lateral pinch data

• Published in: PloS one Vol. 16; no. 9; p. e0255103
• Main Authors: Kearney, Kalyn M, Harley, Joel B, Nichols, Jennifer A
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 02.09.2021; Public Library of Science (PLoS)
• Subjects: Accuracy; Artificial neural networks; Biology and Life Sciences; Biomechanical Phenomena; Biomechanics; Biomedical engineering; Classification; Computer and Information Sciences; Computer Simulation; Datasets; Electromyography - methods; Engineering and Technology; Evaluation; Force; Hand Strength - physiology; Human motion; Humans; Identification and classification; In vivo methods and tests; Long short-term memory; Machine learning; Mathematical models; Medicine and Health Sciences; Model accuracy; Muscle contraction; Muscle, Skeletal - physiology; Muscles; Neural networks; Neural Networks, Computer; Parameter estimation; Pinch force; Research and Analysis Methods; Simulation; Tendons - physiology; Thumb - physiology; United States; Florida; United States > US
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0255103

Hill-type muscle models are widely employed in simulations of human movement. Yet, the parameters underlying these models are difficult or impossible to measure in vivo. Prior studies demonstrate that Hill-type muscle parameters are encoded within dynamometric data. But, a generalizable approach for estimating these parameters from dynamometric data has not been realized. We aimed to leverage musculoskeletal models and artificial neural networks to classify one Hill-type muscle parameter (maximum isometric force) from easily measurable dynamometric data (simulated lateral pinch force). We tested two neural networks (feedforward and long short-term memory) to identify if accounting for dynamic behavior improved accuracy. We generated four datasets via forward dynamics, each with increasing complexity from adjustments to more muscles. Simulations were grouped and evaluated to show how varying the maximum isometric force of thumb muscles affects lateral pinch force. Both neural networks classified these groups from lateral pinch force alone. Both neural networks achieved accuracies above 80% for datasets which varied only the flexor pollicis longus and/or the abductor pollicis longus. The inclusion of muscles with redundant functions dropped model accuracies to below 30%. While both neural networks were consistently more accurate than random guess, the long short-term memory model was not consistently more accurate than the feedforward model. Our investigations demonstrate that artificial neural networks provide an inexpensive, data-driven approach for approximating Hill-type muscle-tendon parameters from easily measurable data. However, muscles of redundant function or of little impact to force production make parameter classification more challenging.