MedicalZooPytorch

Why we included this project

For teams doing medical image segmentation in PyTorch, this library collects several 3D network designs, including UNet, VNet, DenseNet, and ResNet variants, plus data loaders for common benchmark sets like BraTS, iSeg, and MRBrainS18. The loaders matter because wrangling volumetric scans into a usable format often eats more time than the modeling itself. The project started as an MSc thesis, and the included Colab notebooks and manual make it a reasonable entry point for students new to 3D segmentation, while the model zoo gives experienced researchers a quick basis for comparisons and experiments.