Related Experiment Video
Updated: Nov 23, 2025
09:10
Digital Hybrid Model Preparation for Virtual Planning of Reconstructive Dentoalveolar Surgical Procedures
Published on: August 5, 2021
2.0K
TSegNet: An efficient and accurate tooth segmentation network on 3D dental model
Zhiming Cui1, Changjian Li2, Nenglun Chen3
1Department of Computer Science, The University of Hong Kong, Hong Kong, China; School of Biomedical Engineering, ShanghaiTech University, Shanghai, China.
Medical Image Analysis
|January 2, 2021
Summary
TSegNet accurately segments teeth in 3D dental models, even with missing or misaligned teeth. This novel method improves accuracy and speed for computer-aided dentistry applications.
Area of Science:
- Computer-aided dentistry
- 3D imaging and analysis
- Machine learning for medical applications
Background:
- Accurate segmentation of dental models is crucial for computer-aided dentistry.
- Existing methods struggle with challenging clinical cases like missing, crowded, or misaligned teeth.
- Robust tooth segmentation is needed for effective orthodontic treatment planning.
Purpose of the Study:
- To propose TSegNet, a novel end-to-end learning-based method for robust and efficient tooth segmentation.
- To address limitations of previous methods in handling abnormal dental models.
- To improve the accuracy and speed of tooth segmentation on 3D scanned point cloud data.
Main Methods:
- TSegNet utilizes a two-stage approach: distance-aware tooth centroid voting for localization and a confidence-aware cascade segmentation module.
- The method processes 3D scanned point cloud data of dental models.
- Evaluation was performed on a large-scale real-world dataset of pre- and post-orthodontic treatment dental models.
Main Results:
- TSegNet demonstrates robust and accurate tooth labeling in challenging clinical cases.
- The method significantly outperforms state-of-the-art approaches, improving Dice Coefficient by 6.5% and F1 score by 3.0%.
- TSegNet achieves a 20-fold speedup in computational time compared to existing methods.
Conclusions:
- TSegNet offers a robust and efficient solution for tooth segmentation in computer-aided dentistry.
- The proposed method effectively handles complex dental anomalies, improving treatment planning.
- TSegNet represents a significant advancement in automated dental model analysis.
