Caries DETR: Tooth Structure-aware Prior and Lesion-aware Dynamic Loss Refinement for DETR Based Caries Detection

2026-04-26

AI Summary

This work addresses the challenge of detecting early dental caries in intraoral images, where lesions often exhibit low contrast and subtle morphological features. To tackle this, the authors propose a DETR-based framework incorporating two key components: Tooth Structure-aware Query Initialization (TSQI) and Lesion-aware Dynamic Loss Refinement (LDLR). TSQI leverages anatomical priors of tooth structures to guide object queries, while LDLR adaptively reweights the loss function based on lesion characteristics and prediction quality, enabling efficient hard example mining. Integrated within a Transformer detection architecture and enhanced by large-scale pretraining on intraoral images, the proposed method achieves state-of-the-art performance on the AlphaDent and DentalAI datasets, significantly improving detection accuracy for minute carious lesions while demonstrating strong generalization and robustness.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Neural Architectures and Foundation Models
• Search and Optimization: Learning to Search

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

As dental caries appear as subtle, low-contrast lesions in intraoral imaging, existing deep learning models face significant challenges in the early detection of caries. While recent Transformer-based detectors have shown promising results in natural images, they often fail to capture the domain-specific anatomical priors crucial for dental caries detection. In this paper, we propose Caries-DETR, a specialized Transformer framework for caries detection in intraoral images. A Tooth Structure-aware Query Initialization (TSQI) is designed, leveraging large-scale intraoral photograph pre-training and a structure perception branch (SPB) to integrate high-frequency structural priors, guiding the model to focus on anatomically significant lesion areas. Furthermore, we design a Lesion-aware Dynamic Loss Refinement (LDLR) to implement quality-driven hard mining through adaptive loss reweighting based on lesion size, anatomical relevance, and prediction quality, optimizing detection for subtle lesions. Extensive experiments on two public datasets (i.e., AlphaDent and DentalAI) demonstrate that Caries-DETR achieves a state-of-the-art performance compared to existing methods and exhibits good generalization and robustness. Code and data at https://github.com/XuefenLiu-SZU/Caries-DETR}{https://github.com/XuefenLiu-SZU/Caries-DETR.

Problem

• dental caries detection
• intraoral imaging
• subtle lesions
• anatomical priors
• low-contrast lesions

Innovation

• Tooth Structure-aware Query Initialization
• Lesion-aware Dynamic Loss Refinement
• DETR
• Caries Detection
• Intraoral Imaging