yeigen/nih-chest-xray

--- dataset_info: features: - name: png dtype: image splits: - name: train num_examples: 112120 dataset_size: ~45GB task_categories: - image-classification language: - en tags: - medical - chest-xray - radiology - NIH --- # NIH ChestX-ray14 — WebDataset Dataset original: [nih-chest-xrays/data](https://www.kaggle.com/datasets/nih-chest-xrays/data) - **Imágenes:** 112,120 PNGs en escala de grises (1024×1024) - **Formato:** WebDataset (TAR archives, ~1000 imágenes por TAR) - **Split:** `train` ## Cómo cargar ```python from datasets import load_dataset ds = load_dataset("webdataset", data_dir="yeigen/nih-chest-xray", split="train", streaming=True) for sample in ds: img = sample["png"] # PIL Image print(img.size) break ```