Tracking high-speed patella motion using biplanar videoradiography : an accuracy study

Patellofemoral pain (PFP) is one of the most common disorders of the knee. However, the pathomechanics of PFP have not been fully elucidated. Our lack of understanding of PFP is partially due to difficulties with tracking the 3-D kinematics of the patellofemoral joint during important dynamic activities, such as running. Biplanar videoradiography can track dynamic 3-D bone motion with high accuracy, and is a promising modality for elucidating the pathomechanics of PFP. Recently, accuracies of 0.1 mm and 0.15â° were reported when tracking motion of the distal femur, radius, and ulna [1]. However, the accuracy of markerless tracking algorithms inherently depends on bone morphology. Therefore, the purpose of this study was to evaluate the systematic error of a biplanar videoradiography system while tracking dynamic motion of the patella using a markerless bone registration algorithm. Secondarily, we introduce and evaluate a filtering step that directly operates on the 3-D transforms computed from biplanar videoradiography data.