14 / 2021-09-12 23:22:52

Automatic cone-beam CT Segmentation with small samples based on Generative Adversarial Networks and semantic segmentation

Segmentation, Generative Adversarial Networks, Annotation

This paper establishes a method to realize semi-automatic or automatic labeling of multi-dimensional data based on small samples and weak labeling. This method could effectively assist doctors in the segmentation of different tissues in dentistry. Based on the U-net combined with the Generative Adversarial Networks method, segmentation can be realized on multi-dimensional data. It also includes three-dimensional mesh reconstruction of the segmented tissue, smooth the boundary, and the result data can be used as clinical aided diagnostic data, or 3D printing data. The result of segmentation can reflect the structural distribution of different tissues. The results could effectively help the doctors and could also help build a mechanical model based on CBCT.