A parametric derivative method for craniography-based orthodontic diagnosis based on machine learning from 3D CBCT images taken from the natural head position.

Abstract

The invention proposes a method for deriving cephalometric parameters for accurate orthodontic diagnosis based on machine learning. More specifically... This invention proposes a method for deriving cephalometric parameters for orthodontic diagnosis based on machine learning. Specifically, the machine learning algorithm is applied after acquiring 3D cone-beam computed tomography (CBCT) images of the subject taken at a natural head position. The cephalometric reference points can then be quickly and accurately digitized on the 3D CBCT image to derive 13 diagnostic parameters for accurate orthodontic diagnosis.