Synthesizing High-Quality Visual Question Answering from Medical Documents with Generator-Verifier LMMs

TL;DR

This paper introduces MedVLSynther, a framework that synthesizes high-quality medical VQA data from biomedical literature using generator-verifier models, enabling improved training of open-weight LMMs for medical question answering.

Contribution

The authors develop a novel generator-verifier pipeline to create large, high-quality medical VQA datasets from open literature, enhancing model training and evaluation.

Findings

Achieved state-of-the-art accuracy on multiple medical VQA benchmarks.

Generated over 13,000 verified medical VQA questions from open biomedical literature.

Demonstrated the necessity of both generation and verification stages for high-quality data.

Abstract

Large Multimodal Models (LMMs) are increasingly capable of answering medical questions that require joint reasoning over images and text, yet training general medical VQA systems is impeded by the lack of large, openly usable, high-quality corpora. We present MedVLSynther, a rubric-guided generator-verifier framework that synthesizes high-quality multiple-choice VQA items directly from open biomedical literature by conditioning on figures, captions, and in-text references. The generator produces self-contained stems and parallel, mutually exclusive options under a machine-checkable JSON schema; a multi-stage verifier enforces essential gates (self-containment, single correct answer, clinical validity, image-text consistency), awards fine-grained positive points, and penalizes common failure modes before acceptance. Applying this pipeline to PubMed Central yields MedSynVQA: 13,087 audited questions over 14,803 images spanning 13 imaging modalities and 28 anatomical regions. Training open-weight LMMs with reinforcement learning using verifiable rewards improves accuracy across six medical VQA benchmarks, achieving averages of 55.85 (3B) and 58.15 (7B), with up to 77.57 on VQA-RAD and 67.76 on PathVQA, outperforming strong medical LMMs. A Ablations verify that both generation and verification are necessary and that more verified data consistently helps, and a targeted contamination analysis detects no leakage from evaluation suites. By operating entirely on open literature and open-weight models, MedVLSynther offers an auditable, reproducible, and privacy-preserving path to scalable medical VQA training data.