Adapting Foundation Model for Dental Caries Detection with Dual-View Co-Training

2025-08-28

AI Summary

Low contrast and morphological variability of dental caries in panoramic radiographs hinder detection accuracy. To address this, we propose DVCTNet, a dual-view collaborative training framework for caries detection that jointly performs global image screening and tooth-level local detail analysis. DVCTNet introduces a gated cross-visual attention module (GCV-Atten) to dynamically fuse multi-scale features, emulating the clinical diagnostic logic of “localization before discrimination.” Technically, it integrates automated tooth segmentation, vision foundation model–based feature extraction, and region proposal optimization. Evaluated on both public benchmarks and a newly constructed high-fidelity clinical dataset, DVCTNet achieves significant improvements over state-of-the-art methods—particularly in detecting early-stage, subtle carious lesions—demonstrating strong clinical applicability and generalizability.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Distributed Search

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Accurate dental caries detection from panoramic X-rays plays a pivotal role in preventing lesion progression. However, current detection methods often yield suboptimal accuracy due to subtle contrast variations and diverse lesion morphology of dental caries. In this work, inspired by the clinical workflow where dentists systematically combine whole-image screening with detailed tooth-level inspection, we present DVCTNet, a novel Dual-View Co-Training network for accurate dental caries detection. Our DVCTNet starts with employing automated tooth detection to establish two complementary views: a global view from panoramic X-ray images and a local view from cropped tooth images. We then pretrain two vision foundation models separately on the two views. The global-view foundation model serves as the detection backbone, generating region proposals and global features, while the local-view model extracts detailed features from corresponding cropped tooth patches matched by the region proposals. To effectively integrate information from both views, we introduce a Gated Cross-View Attention (GCV-Atten) module that dynamically fuses dual-view features, enhancing the detection pipeline by integrating the fused features back into the detection model for final caries detection. To rigorously evaluate our DVCTNet, we test it on a public dataset and further validate its performance on a newly curated, high-precision dental caries detection dataset, annotated using both intra-oral images and panoramic X-rays for double verification. Experimental results demonstrate DVCTNet's superior performance against existing state-of-the-art (SOTA) methods on both datasets, indicating the clinical applicability of our method. Our code and labeled dataset are available at https://github.com/ShanghaiTech-IMPACT/DVCTNet.

Problem

• Detecting dental caries accurately from panoramic X-rays
• Addressing subtle contrast variations and diverse lesion morphology
• Integrating global and local views for improved detection

Innovation

• Dual-view co-training network for caries detection
• Gated cross-view attention module fuses features
• Pretrained foundation models on global and local views