Caries DETR model enhances dental imaging with tooth structure priors

Researchers have developed Caries-DETR, a new Transformer-based framework specifically designed for detecting dental caries in intraoral images. This model incorporates a Tooth Structure-aware Query Initialization to leverage anatomical priors and a Lesion-aware Dynamic Loss Refinement to optimize detection of subtle lesions. Experiments on public datasets show Caries-DETR achieves state-of-the-art performance and demonstrates strong generalization capabilities. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a specialized Transformer model for improved early detection of dental caries, potentially enhancing diagnostic accuracy in dentistry.

RANK_REASON This is a research paper describing a new model for a specific application.

Read on arXiv cs.CV →

• Caries DETR
• Transformer
• AlphaDent
• DentalAI

• paper
• product