“…However, in term of the complexity of methods, the fluoroscopy based techniques require additional model segmentations from the CT scan (Bey et al, 2008(Bey et al, , 2006Nishinaka et al, 2008) or MRI (Boyer et al, 2008;Massimini et al, 2012;Zhu et al, 2012) that risk increased radiation or cost, whereas the present method was based on a priori geometry knowledge from a relatively large cadaveric database and did not require additional image acquisition for shape reconstruction. Besides, compared to conventional motion capture techniques based on skin markers (Ludewig and Cook, 2002), the accuracy of the present method was substantially improved, as the scalupa and humerus were tracked directly on X-ray images that avoided errors from skin motion relative to the bony structures.…”