lamamkh/SGMRI-VQA
收藏Hugging Face2026-04-17 更新2026-04-26 收录
下载链接：
https://hf-mirror.com/datasets/lamamkh/SGMRI-VQA
下载链接
链接失效反馈官方服务：
资源简介：
---
license: apache-2.0
task_categories:
- visual-question-answering
- image-to-text
tags:
- medical-imaging
- mri
- grounding
- spatial-reasoning
- chain-of-thought
pretty_name: SGMRI-VQA
size_categories:
- 10K<n<100K
---
<p align="center">
<img src="https://huggingface.co/front/assets/huggingface_logo-noborder.svg" width="120"/>
</p>
# SGMRI-VQA: Spatially Grounded MRI Visual Question Answering
A 41,307-pair VQA benchmark for multi-frame spatially grounded reasoning on volumetric MRI, built from expert radiologist annotations in the fastMRI+ dataset across brain and knee studies.
## Contents
- `brain/` — brain MRI QA + metadata (train/val)
- `knee/` — knee MRI QA + metadata (train/val)
- `*_images.tar` — image archives per split
## Paper
*Beyond a Single Slice: Multi-Frame Spatially Grounded Reasoning Across Volumetric MRI*
## Model
Fine-tuned model: [lamamkh/Qwen3-8B-SGMRIVQA-SFT](https://huggingface.co/lamamkh/Qwen3-8B-SGMRIVQA-SFT)
提供机构：
lamamkh
