Confidence Surface-Based Fine Matching Between Dental Cone-Beam Computed Tomography Scan and Optical Surface Scan Data

In orthognathic or dental surgery, cone-beam computed tomography (CBCT) and optical surface scanning are commonly used in procedures for accurate diagnosis and treatment planning. Because the CBCT and optical scan surface have different scan ranges and orientations, surface matching is required to use the both scan data simultaneously. In this study, we use a highly reliable scan surface, called confidence surface, for improving the matching accuracy. To do this, we propose a method to measure the noise possibility, that is, the uncertainty of each surface. This is accomplished by estimation methods of geometric variation with respect to different scan data. An iterative closest point-based fine matching is performed using only the confidences surfaces. The matching accuracy was shown by root-mean-square-error (RMSE) and color-coded visualization of the distance between the surfaces obtained from different scanning data. The average RMSE of matching result by the proposed method was 0.3298. This experimental results show that our proposed method improves the accuracy than the comparison methods, even with CBCT data that includes severe metal artifacts.