English

Synthetic Vasculature and Pathology Enhance Vision-Language Model Reasoning

Computer Vision and Pattern Recognition 2025-12-15 v1

Abstract

Vision-Language Models (VLMs) offer a promising path toward interpretable medical diagnosis by allowing users to ask about clinical explanations alongside predictions and across different modalities. However, training VLMs for detailed reasoning requires large-scale image-text datasets. In many specialized domains, for example in reading Optical Coherence Tomography Angiography (OCTA) images, such precise text with grounded description of pathologies is scarce or even non-existent. To overcome this bottleneck, we introduce Synthetic Vasculature Reasoning (SVR), a framework that controllably synthesizes images and corresponding text, specifically: realistic retinal vasculature with Diabetic Retinopathy (DR) features: capillary dropout, microaneurysms, neovascularization, and tortuosity, while automatically generating granular reasoning texts. Based on this we curate OCTA-100K-SVR, an OCTA image-reasoning dataset with 100,000 pairs. Our experiments show that a general-purpose VLM (Qwen3-VL-8b) trained on the dataset achieves a zero-shot balanced classification accuracy of 89.67% on real OCTA images, outperforming supervised baselines. Through human expert evaluation we also demonstrate that it significantly enhances explanation quality and pathology localization on clinical data.

Keywords

Cite

@article{arxiv.2512.11060,
  title  = {Synthetic Vasculature and Pathology Enhance Vision-Language Model Reasoning},
  author = {Chenjun Li and Cheng Wan and Laurin Lux and Alexander Berger and Richard B. Rosen and Martin J. Menten and Johannes C. Paetzold},
  journal= {arXiv preprint arXiv:2512.11060},
  year   = {2025}
}

Comments

23 pages, 8 figures, 6 tables. Full paper under review for MIDL 2026 (Medical Imaging with Deep Learning)