# IOSVLM: A 3D Vision-Language Model for Unified Dental Diagnosis from Intraoral Scans

Source: [IOSVLM: A 3D Vision-Language Model for Unified Dental Diagnosis from Intraoral Scans](<https://www.sota2.com/research/paper/69ba31b70636e5bf4afc0aad>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-19T05:00:12.344Z

Authors: Huimin Xiong, Zijie Meng, Tianxiang Hu, Chenyi Zhou, Yang Feng, Zuozhu Liu

Published: 2026-03-17T16:57:02.000Z

## Abstract

3D intraoral scans (IOS) are increasingly adopted in routine dentistry due to abundant geometric evidence, and unified multi-disease diagnosis is desirable for clinical documentation and communication. While recent works introduce dental vision-language models (VLMs) to enable unified diagnosis and report generation on 2D images or multi-view images rendered from IOS, they do not fully leverage native 3D geometry. Such work is necessary and also challenging, due to: (i) heterogeneous scan forms and the complex IOS topology, (ii) multi-disease co-occurrence with class imbalance and fine-grained morphological ambiguity, (iii) limited paired 3D IOS-text data. Thus, we present IOSVLM, an end-to-end 3D VLM that represents scans as point clouds and follows a 3D encoder-projector-LLM design for unified diagnosis and generative visual question-answering (VQA), together with IOSVQA, a large-scale multi-source IOS diagnosis VQA dataset comprising 19,002 cases and 249,055 VQA pairs over 23 oral diseases and heterogeneous scan types. To address the distribution gap between color-free IOS data and color-dependent 3D pre-training, we propose a geometry-to-chromatic proxy that stabilizes fine-grained geometric perception and cross-modal alignment. A two-stage curriculum training strategy further enhances robustness. IOSVLM consistently outperforms strong baselines, achieving gains of at least +9.58% macro accuracy and +1.46% macro F1, indicating the effectiveness of direct 3D geometry modeling for IOS-based diagnosis.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Multi-disease IOS diagnosis on IOSVQA (test)](<https://www.sota2.com/research/sota/multi-disease-ios-diagnosis-on-iosvqa-test.md>): Multi-disease IOS diagnosis; IOSVQA (test); Accuracy: 77.23.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.16781>)

[PDF](<https://arxiv.org/pdf/2603.16781.pdf>)
