MSK-Morph: An automated framework to systematically morph landmark-defined musculoskeletal models into subject-specific bone geometries

Abstract

Musculoskeletal models are widely used to study human movement, investigate musculoskeletal disorders and evaluate athletic performance. The accuracy of these models depends primarily on representing subject-specific musculoskeletal geometry, which determines joint definitions and muscle paths. Subject-specific models can be derived from medical imaging, however the task remains labour-intensive with numerous subjective decisions, which limits their reproducibility and use in large-scale studies. Automated methods that preserve anatomical model topology while adapting models to individual bone geometries are therefore needed. Here, we develop and demonstrate a landmark-based morphing framework, MSK-Morph, to systematically transform template musculoskeletal models into subject-specific models based on bone geometry derived from medical imaging. MSK-Morph introduces an anatomical landmark-defined musculoskeletal model that embeds segment and joint definitions, and muscle paths, and uses them to systematically and reproducibly morph the model to target bone geometries. MSK-Morph automatically updates the joint definitions and muscle paths to reflect inter-individual skeletal variation while maintaining the structural topology of the original model. MSK-Morph produces landmark-defined musculoskeletal models that remain compatible with existing simulation workflows. By enabling rapid generation of models with subject-specific skeletal geometry, this framework facilitates large-scale musculoskeletal modelling and the development of more diverse generic model libraries.