Tooth orthodontic result prediction method and system based on feature point recognition

Abstract

The invention discloses a tooth orthodontic result prediction method and system based on feature point identification. The method comprises the following steps: obtaining a dental crown point cloud, a tooth point cloud and a tooth label according to a dentition model; extracting dentition global features and tooth local features according to the dental crown point cloud and the tooth point cloud, performing feature fusion on the dentition global features, the tooth tag of the single tooth and the tooth local features, and extracting feature points of the single tooth according to the fusion features and offset vectors of points and feature points in the tooth point cloud; fusing the feature points of the single tooth with the tooth point cloud to extract tooth attention features, obtaining dentition attention features according to the tooth attention features, and fusing the dentition attention features, the global dentition features and the local tooth features to obtain a point cloud of fused feature points; and obtaining rigid body transformation parameters before and after tooth orthodontics according to the point cloud of the fused feature points and the dental crown point cloud of the target position, and obtaining a dental crown model prediction result after orthodontics according to the rigid body transformation parameters and the dental crown point cloud of the initial position to be orthodontic.

Description

technical field [0001] The invention relates to the technical fields of deep learning and image processing, in particular to a method and system for predicting orthodontic results based on feature point recognition. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] Orthodontics involves the diagnosis, correction and prevention of tooth and jaw misalignment. With the development of 3D printing technology, invisible aligners have been introduced into the field of orthodontics, and at the same time, the research field of digital orthodontics has been developed. After research, there are now many invisible aligner service providers for orthodontists. Most of them are used by doctors as a window to communicate with patients, and the orthodontist gives the orthodontic plan, uploads the patient’s data to the technician for operation, obtains the t...

AI Technical Summary

Problems solved by technology

[0005] However, the diagnosis process greatly depends on the experience level of the orthodontist, and also has strong subjective factors, which has led to many analysis and diagnosis methods that can rely on existing data and markers based on deep learning. Limited to the dependence of deep learning methods on data sets, different medical data collection devices lead to data alignment problems and privacy issues of medical data, making the robustness of deep learning algorithms a question of whether an algorithm can get better results under different data sets The key; therefore, the orthodontic prediction directly using deep learning is still a very challenging work

[0006] For orthodontics, it is far from enough to only have a 3D model of the teeth. The existing automatic orthodontic solution generation directly performs end-to-end output processing on the input data, which lacks interpretability and reliability. Even if the doctor gets It is impossible to judge whether the obtained orthodontic plan is accurate

Images