Modelling Soft-Tissue Motion During Human Movement Experiments to Improve Calculations of Skeletal Kinematics

Abstract

In Canada, approximately 544,000 upper-limb injuries occurred in a 12-month period between 2009 and 2010, many of which were injuries to the rotator cuff muscles of the shoulder. Because of the complex structure and function of the shoulder, it is often difficult to determine which muscles have been injured. The most widely used technology to study human movement is motion capture, wherein markers are affixed to a subject’s skin and are tracked by cameras as the subject moves. The recorded marker trajectories are then used to estimate the bone locations and joint angles during the tracked motion. This is called an inverse kinematic simulation. The simulation can then be used to estimate variables that are difficult or impossible to measure directly, such as the activation of single muscle heads within a muscle group. However, muscles bulge and skin stretches during movement, so the markers that are affixed to the skin generally move relative to the underlying bones. These errors, known as soft-tissue artifacts, lead to uncertainty in the calculation of bone locations and, consequently, uncertainty in the computed skeletal joint angles. This uncertainty limits the use of inverse kinematic simulations in clinical settings. Given the skin tissue’s elastic behaviour, a spring-based equilibrium model can be used to estimate the behaviour of skin during non-impulsive motion. In the proposed model, markers were placed on the surface of ellipsoids (representing the thorax, abdomen, scapula, and upper arm) and were attached to each other via springs. The system was assumed to remain in static equilibrium during sufficiently slow movements to approximate the stretch of the skin. In this thesis, the development and application of a proof-of-concept model to estimate the pose of the skeleton is described. This work demonstrates the feasibility of using such a model to reduce errors due to soft-tissue motion.