Research

Paper

TESTING March 06, 2026

OralGPT-Plus: Learning to Use Visual Tools via Reinforcement Learning for Panoramic X-ray Analysis

Authors

Yuxuan Fan, Jing Hao, Hong Chen, Jiahao Bao, Yihua Shao, Yuci Liang, Kuo Feng Hung, Hao Tang

Abstract

Panoramic dental radiographs require fine-grained spatial reasoning, bilateral symmetry understanding, and multi-step diagnostic verification, yet existing vision-language models operate under a static single-pass paradigm that limits their clinical reliability. In this paper, we introduce OralGPT-Plus, an agentic vision-language model designed to perform iterative and symmetry-aware diagnostic reasoning for panoramic dental radiograph analysis. To support this paradigm, we construct DentalProbe, a five-thousand-image dataset with expert-curated diagnostic trajectories that provide structured supervision for localized inspection and contralateral comparison. We further develop a Reinspection-driven reinforcement learning framework that encourages clinically meaningful re-examination and stabilizes long-horizon reasoning with rubric-based reward and conditioned diagnostic-driven reward. In parallel, we present MMOral-X, the first benchmark for holistic panoramic diagnosis, containing 300 open-ended questions and region-level annotations across multiple difficulty levels. OralGPT-Plus demonstrates consistent and reliable improvements over strong baselines on MMOral-X and established panoramic benchmarks, indicating the effectiveness of interactive and symmetry-informed reasoning. Our work highlights the value of agentic modeling for dental imaging and provides a foundation for future research in clinically aligned panoramic radiograph analysis.

Metadata

arXiv ID: 2603.06366
Provider: ARXIV
Primary Category: cs.CV
Published: 2026-03-06
Fetched: 2026-03-09 06:05

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.06366v1</id>\n    <title>OralGPT-Plus: Learning to Use Visual Tools via Reinforcement Learning for Panoramic X-ray Analysis</title>\n    <updated>2026-03-06T15:16:30Z</updated>\n    <link href='https://arxiv.org/abs/2603.06366v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.06366v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Panoramic dental radiographs require fine-grained spatial reasoning, bilateral symmetry understanding, and multi-step diagnostic verification, yet existing vision-language models operate under a static single-pass paradigm that limits their clinical reliability. In this paper, we introduce OralGPT-Plus, an agentic vision-language model designed to perform iterative and symmetry-aware diagnostic reasoning for panoramic dental radiograph analysis. To support this paradigm, we construct DentalProbe, a five-thousand-image dataset with expert-curated diagnostic trajectories that provide structured supervision for localized inspection and contralateral comparison. We further develop a Reinspection-driven reinforcement learning framework that encourages clinically meaningful re-examination and stabilizes long-horizon reasoning with rubric-based reward and conditioned diagnostic-driven reward. In parallel, we present MMOral-X, the first benchmark for holistic panoramic diagnosis, containing 300 open-ended questions and region-level annotations across multiple difficulty levels. OralGPT-Plus demonstrates consistent and reliable improvements over strong baselines on MMOral-X and established panoramic benchmarks, indicating the effectiveness of interactive and symmetry-informed reasoning. Our work highlights the value of agentic modeling for dental imaging and provides a foundation for future research in clinically aligned panoramic radiograph analysis.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CV'/>\n    <published>2026-03-06T15:16:30Z</published>\n    <arxiv:comment>34 pages, 24 figures, conference</arxiv:comment>\n    <arxiv:primary_category term='cs.CV'/>\n    <arxiv:journal_ref>CVPR 2026</arxiv:journal_ref>\n    <author>\n      <name>Yuxuan Fan</name>\n    </author>\n    <author>\n      <name>Jing Hao</name>\n    </author>\n    <author>\n      <name>Hong Chen</name>\n    </author>\n    <author>\n      <name>Jiahao Bao</name>\n    </author>\n    <author>\n      <name>Yihua Shao</name>\n    </author>\n    <author>\n      <name>Yuci Liang</name>\n    </author>\n    <author>\n      <name>Kuo Feng Hung</name>\n    </author>\n    <author>\n      <name>Hao Tang</name>\n    </author>\n  </entry>"
}