Abstract Summary
The current study aimed to develop a statistical shape model (SSM) for reconstruction of subject-specific three-dimensional (3D) knee model with multiple fluoroscopic images, and to evaluate via computer simulation its theoretical accuracy and determine the minimum number of images needed for an acceptable accuracy. The results showed that reconstruction with six views had the highest accuracy, suggesting that the SSM method has sufficient accuracy for future application in joint kinematics measurement using 3D fluoroscopy with minimum radiation.