gerstung-lab / wbs-prediction

Code repository for a project focused on diagnostic prediction from whole blood slides

A complete computational assessment of the cytomorphological determinants of myelodyplastic syndromes

Motivation

This is the repository for PLACEHOLDER. In this work, we use the whole blood slides of >300 individuals with myelodyplastic syndromes and anaemias and use them to develop a method that is capable of predicting a disease and retrieving examples of cells which are relevant for each classification.

Code map

Requirements

Software

• python
• snakemake
• R (analysis and plotting)

Required python packages

opencv-python, tensorflow==1.12, scikit-image, h5py, albumentations, psutil, pytorch, tifffile

Project enumeration

1. pipeline - contains the pipeline for WBC and RBC detection and characterisation from WBS
2. simulations - contains simulations validating MILe-ViCe
3. mile-vice - contains the code to train and run MILe-ViCe on the output from pipeline
4. rbc-segmentation - contains the code to train a predictor that filters poorly predictions for detected RBC
5. (STILL TESTING) vae-characterisation - characterisation of blood cells using a beta-variational autoencoder