---
license: apache-2.0
task_categories:
  - visual-question-answering
  - image-to-text
language:
  - en
tags:
  - medical
  - radiology
  - vision-language
---

# Medical Vision-Language Dataset

Combined dataset for training medical vision-language models.

## Statistics
- **Total**: 3793 samples
- **Train**: 3035 | **Validation**: 758

## Sources
- ROCO: 2000
- VQA-RAD: 1793

Created: 2025-11-23 22:25:36