Skip to content

Dataset · RadGenome

RadGenome-ChestCT: Chest CT with texts

RadGenome/RadGenome-ChestCT

Large grounded vision-language dataset for chest CT analysis; metadata lists text, 100K–1M.

What's inside

The card describes RadGenome-ChestCT as a grounded vision-language dataset for chest CT analysis. It is designed to advance medical foundation models and includes diverse supervision signals. The metadata lists the modality as text and the size from 100K to 1M examples; no other details on language or format are provided in the excerpt.

What you can do with it

  • Training a medical vision-language model on chest CT
  • Evaluating grounding of text descriptions on CT
  • Fine-tuning a radiology report generation model
Size
100K–1M rows
Downloads this month
36.8K
Likes
70

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("RadGenome/RadGenome-ChestCT")