Multimodal medical imaging (CT and dynamic MRI) data and computer-graphics multi-physical model for the estimation of patient specific lumbar spine muscle forces

Abstract Computer-graphics multi-physical model has been used to assist the clinician in their decision-making processes. In particular, patient specific musculoskeletal modeling using medical imaging data and physical laws has demonstrated great potential for future clinical analysis of the lumbar spine. The main objective of this present work was to propose a data-driven modeling workflow to create computer-graphics multi-physical model from multimodal medical imaging data to extract useful clinical simulation knowledge leading to better diagnosis and treatment of human diseases such as low back pain. Computed Tomography (CT) data and tissue-based physical laws were used to create geometries as well as to compute full patient specific anthropometrical properties of a patient specific multi-physical lumbar spine model. Kinematical range of motion and spinal curvatures were derived from in vivo dynamic MRI. Then, these multimodal data were combined into the developed model to estimate the lumbar spine muscle forces using inverse dynamics and static optimization. Finally, kinematic behavior of the developed model was evaluated. As results, maximal estimated forces of all muscle groups range from 3 to 40 N for hyperlordosis motion. The higher muscle forces were estimated in iliocostalis lumborum pars lumborum muscle group. The simulated spinal curvatures ranging from 2.7909 to 3.1745 (1/m) are within the range of values (from 2.02 to 9.6142 (1/m)) measured from in vivo dynamic MRI. This study suggested that multimodal medical imaging data derived from CT and dynamic MRI could be of great interest in the development of computer-graphics multi-physical model as well as in the estimation of kinematical ranges of motion, their evaluation and muscle forces for biomechanical applications.