Abstract
Introduction
Conventional hip radiographs allow surgeons, during preoperative planning, to make important decisions. Size and location of implants are routinely measured by overlaying schematics of the implanted components onto preoperative radiographs. Most currently available planning tools are in two-dimensions (2D), using X-ray images and 2D templates of the implants. Determination of the ideal component size requires two radiographic views of the femur: the anterior-posterior (AP) and the lateral direction. The surgeon uses this information to determine component sizes. Even though this approach has been used for many years leading to very good results, this manual process potentially carries multiple shortcomings. The biggest issue with the AP X-ray image is the fact that it is 2D in nature while the measurement's objective is to obtain three-dimensional (3D) parameters.
Objective
The objective of this study is to derive a methodology to automatically select correct THA implant sizes while keeping the anatomical center of each specific patient within a forward solution model (FSM) that predicts post-operative outcomes.
Methods
The femoral components in our process contain five parameters: stem length, neck offset, neck length, neck shaft angle, and component width. There are many steps to measure the morphologic parameters of a femoral component. (1)Preparation of training implant database, (2)defining multi-plane intersection, (3)determining circumcircles for all intersected femoral component contours, (4)finding centers and radii of circumcircles, (5)measuring distances from each circumcircle to the femoral component head center, and (6)determining the stem shaft axis. The FSM fits specific femoral canal using a 3D mesh model of the femur. The femoral component and canal morphology of a femur model are compared to the training femoral component database. For each femoral component morphology, the algorithm determines how far distally the femoral component fits within the canal before collision between the stem and cortical bone. Once the defined position is confirmed, the relative distance from the anatomical femoral head center to the femoral component head center is calculated. This process is repeated for all femoral component morphology. The best fitting femoral component is determined when the distance from its head center to the femoral head center is minimized, Figure 1.
Results
Three intensive validation tools have been developed: (1) cross-sectional analysis, (2) slice analysis, and (3) contact map analysis. Cross-sectional analysis is a graphic interaction program where users can freely view the anatomy at any orientation, Figure 2. The slice analysis enhances the user visualization by providing a static view of the fit between chosen femoral component and femoral canal, Figure 3. Finally, the contact map analysis allows for visualization of contact area through the bone-stem interface.
Conclusion and Discussion
This is a powerful tool with the FSM that allows surgeons to get a “best fit” implant in 3D, based on canal fit and distance from anatomical femoral head center. Surgeons may want to manually size up or down, but the program will pick best fit sizes based on anatomical morphology. Future iterations will consider the reaming depth each surgeon uses to improve implant selection for each surgeon's technique.
For any figures or tables, please contact authors directly.
