The mutual interactions of muscles during exercise is one of the most difficult images to visualize. These interactions are usually only estimated by observations made from the skin surface. We aimed to develop a technique to visualize muscle dynamics with quantitative 3D imaging. In this system the skeletal system and muscle system were reconstructed from 3D data sets obtained by CT and MRI.
　The skeletal system was reconstructed from CT 3D data sets and each joint could be moved using the location data obtained by motion capture system in 3D. In the muscle system, the shape and size of each muscle was obtained from MRI 3D data sets measured from the same subject that the skeletal structure was obtained from. We added a few features for this 3D muscle model. These muscle models are composed of elements of muscle fibers and each element is able to contract or relax like real muscle. In Addition, these muscle fibers are able to contract and relax by avoiding the bone or adjacent muscle fibers as real muscle acts in the body by accounting for physical interference. These components of virtual muscle fibers are used to form a 3D structure by fitting the shape and volume according to the reconstructed 3D muscle image of the subject. Finally, each muscle is attached to the skeletal system one by one in its anatomical position. According to the movements of the skeletal system measured by motion capture system, muscle model could be manipulated. In this system, it is possible to record and visualize the dynamics of each muscle and their relative interactions in space and in time sequential domains quantitatively. This system should provide a useful tool to analyze muscle dynamics anatomically and physiologically in 4D space.

DESCENTE SPORTS SCIENCE Vol.20/THE DESCENTE AND ISHIMOTO MEMORIAL FOUNDATION FOR THE PROMOTION SPORTS SCIENCE