Dataset · ibrahimhamamci
CT scans with reports for 3D vision-language models
ibrahimhamamci/CT-RATE
A set of CT scans with paired radiology reports for training and evaluating 3D VLMs.
What's inside
Contains a multimodal medical dataset: CT images and associated radiology reports. Based on these data, the team is organizing the VLM3D challenge; MR-RATE, a large brain MRI dataset with reports, is also under development. Language is English, size is from 10k to 100k examples.
What you can do with it
- Training 3D VLMs for CT analysis
- Zero-shot classification and VQA on medical images
- Evaluating models on the image-to-text task
- Tasks
- Image → text, Text → image, Image classification, Question answering, Visual question answering, Zero shot classification
- Languages
- English
- Size
- 10K–100K rows
- Downloads this month
- 154.2K
- Likes
- 413
- Access
- on request on Hugging Face
Models trained on it
- nvidia/NV-Reason-CTImage text → text 4.1K
- Kentucky-Open-Science/DALE-CT-1SImage feature extraction 202
- yinghemedical/U-VLMImage → text 189
- Kentucky-Open-Science/DALE-CT-0Image feature extraction 160
- yinghemedical/VLM3D2025-Task1-ReportGenerationImage → text 145
- Kentucky-Open-Science/DALE-CT-2SImage feature extraction 144
- hcltech-robotics/rev-ctImage classification 105
- Deepnoid/GLINT-CTEmbeddings 80
How to load it
from datasets import load_dataset
ds = load_dataset("ibrahimhamamci/CT-RATE")