melisklc0/stroke-classification

Stroke Classification: Pre-processed Brain CT Dataset Related resources Live demo Training code Distilled model Summary Task: Binary CT image classification (No-Stroke = 0, Stroke = 1) Dataset Size: ~16.5k images per fold, ~49.5k files across folds on disk. ~3 GB total Structure: 3-fold cross-validation (Fold1, Fold2, Fold3), each containing train/ and test/ splits Labeling Strategy: Original classes (No Stroke, Bleeding, Ischemia) were… See the full description on the dataset page: https://huggingface.co/datasets/melisklc0/stroke-classification.