################################################################################################ This repository contains code and a description of the data preprocessing pipeline for our work

"How to learn from unlabeled volume data: Self-Supervised 3D Context Feature Learning" published at the MICCAI 2019 conference in Shenzhen, China.

################################################################################################

Find our paper here: <TODO: insert springer URL>

The overall process basically contains 3 major blocks:

1. preprocessing the VISCERAL data set
2. training of several CNN architectures
3. evaluating the image descriptors by organ labeling using an approximate kNN-search

For the first and last step, we provide detailled descriptions in the according txt-files. The training procedures for our proposed feature extractor CNNs is provided in form of a jupyter notebook.