Automatic determination of patellofemoral kinematic parameters through X-ray image processing

Examination of plain Merchant-view X-rays of the patellofemoral joint is often used to diagnose patellofemoral joint malalignment and to qualify the pathology. As digital image processing becomes more widespread, an image processing algorithm is being developed for medical personnel to easily display, manipulate and analyze digitized plain Merchant-view X-ray films on personal computers. The method provides measurement of the kinematic parameters: sulcus angle (SA), patellofemoral congruence angle (PFC), lateral patellar angle (LPA), patellar rotation angle (PR), and lateral patellar displacement (LPD). Smooth edges of the patella and the femur are obtained by using image enhancement, edge detection and morphological thinning techniques based on the thresholds chosen manually for boundary detection. Six landmarks for determining the kinematic parameters are marked automatically on the edges of patella and femur. Parameter measurements for twelve X-ray films of six patients are calculated for analysis and comparison with hand calculated parameters. The goal is develop an algorithm that can evaluate the geometry of the patellofemoral joint, and is cost effective.