Neural Image Compression for Gigapixel Histopathology Image Analysis

This repository contains links to code and data supporting the experiments described in the following paper:

D. Tellez, G. Litjens, J. van der Laak and F. Ciompi
Neural Image Compression for Gigapixel Histopathology Image Analysis
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019
DOI: 10.1109/TPAMI.2019.2936841

The paper can be accessed in the following link: https://doi.org/10.1109/TPAMI.2019.2936841

To create a synthetic dataset use synthetic_data_generation.py or directly downloaded from https://doi.org/10.5281/zenodo.3381498.

Compress a given whole-slide image. A whole-slide image can be compressed using code in the present repository (featurize_wsi.py) and pretrained models (./models/encoders_patches_pathology/*.h5). Requires first vectorizing a slide with vectorize_wsi.py

To compress patches, see featurize_patch_example.py

You can also use https://grand-challenge.org to featurize whole slides via run_nic_gc.py. For this you need an account capable of running algorithms and a token. Contact the administrators for gaining access to these features.

Requirements: keras 2.2.4 and tensorflow 1.14 SimpleITK for converting the grandchallenge-created features to npy.