The wrist is a complex joint with complex dynamics that have thus far eluded a complete understanding. We are trying to model the wrist computationally. An understanding of the kinematics of the wrist has the potential to impact treatment for a number of debilitating conditions that affect millions of people in the U.S..|
We propose to extract models of the bones of the wrist from CT volume images. Currently, the model extraction process involves a labor-intensive manual segmentation for each bone in each of eight to ten volume images per subject. We are trying to automate the segmentation initially so that it need only be done once per subject. We intend to further automate the remaining manual segmentation step, or at least to significantly reduce manual intervention.
Our further interests in the models are to simulate wrist motion. At present, there is a paucity of 3-D kinematic data available. We intend to detail the complete 3-D in-vivo kinematics of the wrist. This knowledge will enhance understanding of normal wrist function and will lead to additional studies on cartilaginous and ligamentous tissues.
We use a signed-distance representation for the bones to implement constraints that keep the ligaments that hold the bones together from passing through the bones. Simulations provide us with ligament length measurements as a function of hand pose. We examine the ligament lengths to understand joint limits and functional deficits for some subjects. We plan to ultimately use the signed-distance volumes to represent the thin layer of cartilage on the surfaces of joints where bones contact.
Our further goals are to determine the effects of various pathologies on the kinematics of the wrist, with the aim of providing improved diagnostic and treatment guidelines for clinical care.
The impact of this work is expected to be both in computational tools that can be used to predictively model and understand these and other joints and an understanding of the anatomy and dynamics of the wrist. This understanding may have applications in biology, bioengineering, and medical applications. Potential future applications to animation and robotics are also likely. The numerical methods we are developing for simulating joints are expected to apply to simulation of other biological systems.