galilai-group/medianomaly

MedIAnomaly is a benchmark for evaluating anomaly detection methods on seven diverse medical imaging datasets: RSNA, VinCXR, BrainTumor, LAG, ISIC2018_Task3, Camelyon16, and BraTS2021. It supports both image-level classification and pixel-level segmentation tasks. All datasets follow a consistent one-class learning protocol: the training set contains only normal (non-anomalous) images, while the test set includes both normal and abnormal cases. This setting is designed to reflect real-world scenarios where anomalous samples are rare or unavailable during training. MedIAnomaly provides standard

<iframe src="https://silytics.si/embed/item/7d102f95-160b-4f38-8c23-3da1971d9277" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.