Image based methods to generate subject-specific musculoskeletal models for gait analysis

This paper describes a workflow, designed to build detailed, subject-specific musculoskeletal models using magnetic resonance (MR) images. Integration of semi-automated methods for segmentation of the bone reduced considerably the processing time. Furthermore, enhanced model quality was achieved compared to pure manual segmentation techniques. Two different semi-automated segmentation methods were tested and compared. The workflow was successfully applied to MR images of the lower extremity of a non-pathological adult subject: the femur, tibia and fibula were segmented along with lines of action of 25 muscles.