Doctoral Dissertation Atlas-based Automated Surgical Planning for Total Hip Arthroplasty

This paper describes a 3-dimensional (3D) automated preoperative planning method of femoral component (stem) for total hip arthroplasty (THA). The stem planning is formulated as a optimization problem which determines implant parameters such as size, position, and rotation. An objective function is defined using the fitness between the femoral canal and the stem surface. In addition to the objective function, constraints of position and orientation derived from conventional studies of femoral anatomy is also defined. Since the objective function consists of several parameters, a parameter tuning was performed in advance of the performance evaluation. Optimization of the objective function was performed using the Powell method. The proposed method was applied to 17 cases and the errors of the planning results between the automated planning results and the experienced surgeon’s plans were compared. Stem size error was within one size in all cases. Mean positional error and mean orientation error were 3.4 mm and 5.0 deg., respectively. According to a qualitative evaluation performed by an experienced surgeon, 14 cases out of 17 cases were accepted and three cases were rejected. The surgeon commented that the rejected three cases can be accepted after some minor modification. Therefore, the author consider the proposed method would be useful as an automated stem planning method.