A semi-automated 3D kinematic estimation of total knee arthroplasty using x-ray fluoroscopic images

Abstract

To achieve 3D kinematic analysis of total knee arthroplasty (TKA), 2D/3D registration techniques, which use X-ray fluoroscopic images and computer aided design model of the knee implants, have been applied to clinical cases. However, in previous study, spurious edges and noises from the edge detection of the implant silhouette were erased manually, and initial pose setting for each X-ray image also needed intensive manual operations. It has been a serious problem for clinical application. This study presents a semi-automated estimation method for 3D kinematic analysis of TKA using X-ray fluoroscopic images, which includes techniques of the robust estimation and initial pose estimation based on data interpolation. In order to ensure the validity of the proposed method, in vivo experiments were performed. The results showed that the estimation accuracy and stability was sufficient in clinical application (the RMS error: 1 mm, 1 degree, and automation rate: 76 %), and the feasibility and effectiveness of the proposed method was demonstrated.