Image based musculoskeletal modeling allows personalized biomechanical analysis of gait

This paper describes a workflow, for building detailed, subject-specific musculoskeletal models from magnetic resonance (MR) images allowing enhanced biomechanical analysis of gait. Bones are segmented semi-automatically using a hybrid approach while muscles attachments are retrieved automatically by atlas-based non-rigid registration followed by optional interactive correction using a user-friendly interface. Compared to previously proposed methods for MR based musculoskeletal modeling, integration of automated image processing procedures and problem-tailored visualization techniques result in a considerable reduction of the processing time, thus making MR-based musculoskeletal modeling practically feasible and more attractive.status: publishe