A benchmark multimodal oro-dental dataset for large vision-language models

TL;DR

This paper introduces a large, annotated multimodal oro-dental dataset with images, radiographs, and textual records, and demonstrates its effectiveness by fine-tuning vision-language models for dental diagnosis and report generation.

Contribution

It provides a comprehensive, publicly available dataset for AI in dentistry and shows how fine-tuning large models improves diagnostic and reporting tasks.

Findings

Fine-tuned models outperform baselines in anomaly classification.

Models achieve significant improvements in diagnostic report generation.

Dataset enables effective training of vision-language models for dental applications.

Abstract

The advancement of artificial intelligence in oral healthcare relies on the availability of large-scale multimodal datasets that capture the complexity of clinical practice. In this paper, we present a comprehensive multimodal dataset, comprising 8775 dental checkups from 4800 patients collected over eight years (2018-2025), with patients ranging from 10 to 90 years of age. The dataset includes 50000 intraoral images, 8056 radiographs, and detailed textual records, including diagnoses, treatment plans, and follow-up notes. The data were collected under standard ethical guidelines and annotated for benchmarking. To demonstrate its utility, we fine-tuned state-of-the-art large vision-language models, Qwen-VL 3B and 7B, and evaluated them on two tasks: classification of six oro-dental anomalies and generation of complete diagnostic reports from multimodal inputs. We compared the fine-tuned models with their base counterparts and GPT-4o. The fine-tuned models achieved substantial gains over these baselines, validating the dataset and underscoring its effectiveness in advancing AI-driven oro-dental healthcare solutions. The dataset is publicly available, providing an essential resource for future research in AI dentistry.