Video analysis of gait for diagnosing movement disorders

This paper describes a video analysis system, free of markers and setup pro cedures, that quantitatively identified gait abnormalities in real time fro m standard video images. A novel color three-dimensional body model was siz ed and texture mapped to the exact characteristics of a person from video i mages. The kinematics of the body model was represented by a transformation tree to track the position and orientation of a person relative to the cam era. Joint angles were used to track the location and orientation of each b ody part, with the range of joint angles being constrained by associating d egrees of freedom with each joint To stabilize tracking, the joint angles w ere estimated for the next frame. The calculation of joint angles, for the next frame, was cast as an estimation problem, which was solved using an it erated extended Kalman filter. Patients with dopa-responsive Parkinsonism, and age-matched normals, were video taped during several gait cycles with w alking movements successfully tracked and classified. The results suggested that this approach has the potential to guide clinicians on the relative s ensitivity of specific postural/gait features in diagnosis. (C) 2000 SPIE a nd IS&T. [S1017-9909(00)00701-7].