A lot of elderly people need to care for joint disease, and one of the treatments for ankle osteoarthritis is artificial joint replacement surgery. Estimating the relative positional difference (alignment) between the bone before the deformation and the deformed bone, artificial joint replacement surgery can be performed precisely. By using the estimated alignment, the artificial ankle joint can obtain high satisfaction with less pain and enough function as an alternative. Although the bone alignment is currently estimated from X-rays and CT images, it is difficult to measure three-dimensional relationships because it is a two-dimensional image. We evaluate a 3D bone alignment based on Principal Components Analysis (PCA) using image deformation. In the proposed method, the PCA is used to create a three-dimensional bone model based on bone thickness, the bone alignment is estimated by Go-ICP algorithm. In this study, the images of the foot were captured using a CT device and the three-dimensional bone model is created by stacking the CT images. For improving the accuracy of overlap, a reference model is created based on bone thickness using the PCA. then, the bone is overlapped using the Go-ICP algorithm and estimated alignment. Using the proposed method, the amount of the displacement in the Z-axis showed a significant difference in Stage 3B compared to healthy subjects. In addition, it also showed a significant difference in Stage 4 compared to healthy subjects. The amount of the rotation in the Z-axis direction showed a significant difference in Stage 3B compared to healthy subjects.