Shape-preserving Tooth Segmentation from CBCT Images Using Deep Learning with Semantic and Shape Awareness

2025-11-20

AI Summary

Severe tooth shape distortion caused by inter-tooth adhesion in CBCT images significantly degrades segmentation accuracy. Method: We propose a semantic- and shape-aware deep learning framework comprising two key components: (i) target-tooth centroid-guided multi-label learning to explicitly model inter-tooth semantic dependencies, and (ii) a differentiable morphological constraint module that jointly optimizes segmentation outputs and anatomical structure fidelity. Contribution/Results: By integrating multi-task learning, semantic relationship modeling, and explicit shape priors, our method achieves state-of-the-art performance on both internal and external CBCT datasets. Quantitative and qualitative evaluations demonstrate superior boundary localization accuracy, effective mitigation of shape distortion, and enhanced anatomical plausibility of segmentation results compared to existing approaches.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Background:Accurate tooth segmentation from cone beam computed tomography (CBCT) images is crucial for digital dentistry but remains challenging in cases of interdental adhesions, which cause severe anatomical shape distortion. Methods: To address this, we propose a deep learning framework that integrates semantic and shape awareness for shape-preserving segmentation. Our method introduces a target-tooth-centroid prompted multi-label learning strategy to model semantic relationships between teeth, reducing shape ambiguity. Additionally, a tooth-shape-aware learning mechanism explicitly enforces morphological constraints to preserve boundary integrity. These components are unified via multi-task learning, jointly optimizing segmentation and shape preservation. Results: Extensive evaluations on internal and external datasets demonstrate that our approach significantly outperforms existing methods. Conclusions: Our approach effectively mitigates shape distortions and providing anatomically faithful tooth boundaries.

Problem

• Addresses tooth segmentation challenges from CBCT images with interdental adhesions
• Reduces shape ambiguity in dental segmentation using semantic relationship modeling
• Preserves anatomical boundary integrity against severe shape distortion

Innovation

• Deep learning integrates semantic and shape awareness
• Target-tooth-centroid prompts multi-label learning strategy
• Tooth-shape-aware mechanism enforces morphological constraints