AgenTeeth reduces errors in dental X-ray language models
AgenTeeth: A Model-Agnostic Framework for Suppressing Hallucination in Frozen Vision-Language Models on Dental X-Rays via Tool Evidence Injection
Computer Vision and Pattern Recognition
Summary
Dental X-ray images are hard for computer programs that understand both images and text, often leading them to guess based on common patterns rather than what’s actually in the image. The authors created AgenTeeth, a framework that improves these programs by adding specialized tools that detect specific teeth and parts of the mouth. These tools provide clear evidence to the language model, helping it answer questions more accurately without changing the original model. Testing showed that AgenTeeth greatly improved performance on dental imaging questions compared to existing methods.
What this means in practice
- •For dental imaging software developers: Integrate AgenTeeth to improve accuracy of dental X-ray analysis by grounding language models with expert vision tools without retraining.
- •For medical imaging platform teams: Enhance interpretation of panoramic radiographs by incorporating specialized detection tools that provide visual evidence to language models.
Authors
Ahmed Rafid, Fariya Ahmed, Rumman Adib, Mehedi Ahamed, Ajwad Abrar, Tareque Mohmud Chowdhury
Abstract
Vision-language models (VLMs) remain largely unreliable on panoramic dental radiographs and can rely on learned anatomical priors rather than evidence in the image. This is particularly problematic for tooth localization and spatial reasoning, and fine-tuned dental VLMs can retain the same spatial biases. We present AgenTeeth, a model-agnostic, tool-augmented framework that grounds frozen VLMs using seven specialized dental vision experts. A question-aware orchestrator selects the relevant tools, whose detections are mapped to FDI tooth numbers or anatomical regions and returned as structured findings together with annotated image overlays. A fresh synthesis call then answers the question using this evidence, without fine-tuning the underlying VLM. On MMOral-OPG-Bench, AgenTeeth improves four backbone VLMs by 12.9-23.0 percentage points over their baselines. Our strongest configuration reaches 65.66% on open-ended VQA, compared with 45.35% for OralGPT-Plus. The advantage also holds at matched scale: a frozen Qwen2.5-VL-7B-Instruct with AgenTeeth reaches 48.11%, exceeding OralGPT-Plus built on the same backbone after supervised fine-tuning and reinforcement learning for tool use. We release the framework, all seven expert models, and a dentist-annotated dataset for alveolar bone-loss detection in panoramic radiographs.