Prediction of in-vivo Knee Joint Kinematics using a Statistical Shape Modeling

Introduction: Using computed tomography (CT) or magnetic resonance (MR) images to construct 3D knee models has been widely used in biomedical engineering research. Statistical shape modeling (SSM) method is an alternative way to provide a fast, cost-efficient, and subject-specific knee modeling technique. Further, 3D models can be used with a Dual-Fluoroscopic Imaging System (DFIS) to study joint kinematics [1]. This study aimed to evaluate the feasibility of using a combined SSM and DFIS technique for investigation of in-vivo knee kinematics. Methods: A total of 152 knees from healthy subjects was used to build bony surface database of the knee, including distal femur and proximal tibia. Each knee was CT-scanned and modelled using a solid modeling program (Rhinoceros, Robert McNeel & Associates, Seattle, WA). A validated dual-fluoroscopic imaging system (BV Pulsera; Philips, Bothell, WA) was used for subjectspecific SSM creation and in vivo 3D knee joint motion acquisition for two subjects. The fluoroscopic images at full extension were used together with the SSM database to predict a 3D SSM surface model of the knee[2]. The stance phase of gait cycle was performed on a treadmill and captured using the dual-fluoroscopic imaging system. After segmentation of the bony outlines on the fluoroscopic images, both CT and SSM model was imported into the virtual fluoroscopic imaging environment for 2D-3D registration. A feature-based 2D-3D automated registration algorithm was used to reproduce the femur and tibia poses using both the CT model and the SSM model [3]. The 2D-3D image-model registration procedure was accomplished through a minimization of the distances between the bony contours of the knee on the fluoroscopic images and projections of the 3D models [2]. The kinematics data predicted using the SSM model was compared to those predicted using the CT model. Results: During the treadmill gait, both the CT model and the SSM model predicted similar knee motion patterns throughout the stance phase (Fig.1 and Fig.2). The SSM model predicted the knee kinematics with average differences of -0.7±1.0 mm for translation and 0.6±1.4° for rotation when compared to the references obtained using the CT model (Table 1).