Inclusion of image-based in vivo experimental data into the Hill-type muscle model affects the estimation of individual force-sharing strategies during walking

Abstract

The study of muscle coordination requires knowledge of the force produced by individual muscles, which can be estimated using Hill-type models. Predicted forces from Hill-type models are sensitive to the muscle’s maximal force-generating capacity (Fmax), however, to our knowledge, no study has investigated the effect of different Fmax personalization methods on predicted muscle forces. The aim of this study was to determine the influence of two personalization methods on predicted force-sharing strategies between the human gastrocnemii during walking. Twelve participants performed a walking protocol where we estimated muscle activation using surface electromyography and fascicle length, velocity, and pennation angle using B-mode ultrasound to inform the Hill-type model. Fmax was determined using either a scaling method or experimental method. The scaling method used anthropometric scaling to determine both muscle volume and fiber length, which were used to estimate the Fmax of the gastrocnemius medialis and lateralis. The experimental method used muscle volume and fascicle length obtained from magnetic resonance imaging and diffusion tensor imaging, respectively. We found that the scaling and the experimental method predicted similar gastrocnemii force-sharing strategies at the group level (mean over the participants). However, substantial differences between methods in predicted force-sharing strategies was apparent for some participants revealing the limited ability of the scaling method to predict force-sharing strategies at the level of individual participants. Further personalization of muscle models using in vivo experimental data from imaging techniques is therefore likely important when using force predictions to inform the diagnosis and management of neurological and orthopedic conditions.