Skip to content

Segmentation of nuclei in histopathology images of kidney tissues using UNet and UNetPPL

Notifications You must be signed in to change notification settings

rshwndsz/nuclei-segmentation

Repository files navigation

Segmentation of H&E stained histopathology images

Getting Started

Clone the repo

git clone https://github.com/rshwndsz/nuclei-segmentation.git nuclei-segmentation

Create a new conda environment and install required packages.

cd nuclei-segmentation
conda env create -f env_<os_name>.yml

Download the dataset into datasets/

Training & Validation

Train the model, with validation in between epochs, on the train_set.

python main.py --phase train

To load model from a previously saved checkpoint use:

python main.py --phase train --load

Testing

Takes a random image from the test_set and tests the performance of the model.

python main.py --phase test

About

Segmentation of nuclei in histopathology images of kidney tissues using UNet and UNetPPL

Topics

Resources

Stars

Watchers

Forks

Languages