Lemonzhoumeng/uda-medical-image-segmentation-sfs

Implementation of Privacy Preserving Domain Adaptation for Semantic Segmentation of Medical Images

Requirements

Python 3.8; Tensorflow 2.2.0; CUDA 460 driver

Data

The MMWHS data can be downloaded from the PnP-AdaNet repository here. The CHAOS data can be downloaded directly from the challenge website. Alternatively, processed data is also available at this link. Data should be placed in a data folder in the repository root.

Running the code

If data pre-processing is needed, please refer to the data-preprocessing-* notebooks. Once data is processed, to run the code simply run training and then adaptation .ipynb notebooks for either problem.