Tooth position prediction method based on adaptive geometry optimization

Abstract

A multi-layer feature optimization Transformer-based tooth position prediction method is proposed to address the problems of difficult access to high-precision medical data and the difficulty of capturing and representing hierarchical features and spatial relationships among teeth by current methods. First, a geometric adaptive optimization strategy and a physiological adaptive reconstruction strategy are designed for real-time adaptation to the complexity of different clinical environments and enhanced pose invariance by integrating the physiological characteristics and anatomical structure of teeth. Then, a hierarchical feature tooth position prediction network was designed to solve the problems of weak ability of MLPs to process high-dimensional data and low accuracy of prediction transformation matrix by extracting hierarchical geometric features of teeth. Finally, a jointly supervised loss function is constructed, which can simultaneously capture the intrinsic differences, spatial relationships and uncertainties of the tooth position prediction disorder distribution, and can effectively supervise the tooth spatial structure relationships and prevent tooth collisions and misalignments. The experimental results show that the accuracy of the proposed method is improved by 2.87% and the rotation and translation errors are reduced by 28.28% and 37.53%, respectively, compared with the current method.

Fig 1. Effect of GAOS strategy to generate dataset.

The top row represents the original dataset, while the bottom row illustrates the visualization effect of GAOS.

Fig 2. Effect of PARS strategy to generate dataset.

The first column represents the original dataset, while the second and third columns illustrate the visualization effects of the PARS strategy.

Fig 3. Overall block diagram of tooth position prediction.

The module on the left is the Data Input Module, the module in the middle is the Hierarchical Feature Extraction Module, and the module on the right is the Ideal Tooth Posture Prediction Module.

Fig 4. Sketch of the key points of the tooth.

The points in different colors represent the key points for each type of tooth.

Fig 5. F-VAMP multi-level tooth position prediction network model.

On the left, the input tooth data and the corresponding features are shown. After processing through the network layers, the output on the right represents the rotation and translation parameters for the ideal tooth position.

Fig 6. Intertooth collision diagram.

Figure (a) shows the tooth collision diagram, while Figure (b) illustrates the effect of resolving the tooth collision issue.

Fig 7. Comparison of tooth position prediction results.

The regions highlighted by the rectangular boxes indicate the orthodontic areas.

Fig 8. Comparison of tooth collision and tooth misalignment.

The regions highlighted by the rectangular boxes indicate the areas where tooth misalignment and tooth collisions have been addressed.