Skip to content
Training FCNNs from patches to full-sized images. A framework to train arbitrarily designed networks for medical image segmentation.
Jupyter Notebook Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.ipynb_checkpoints
datasets Initial Commit Mar 18, 2019
images commenting Apr 8, 2019
.gitignore ignore Apr 8, 2019
LICENSE Initial commit Mar 18, 2019
README.md Update README.md Jul 3, 2019
networks.py
patch_manipulations.ipynb commenting Apr 8, 2019
plotting.py Initial Commit Mar 18, 2019
routines.py Bug fix Mar 18, 2019
seg_101_image_based.ipynb Initial Commit Mar 18, 2019
seg_102_patch_based.ipynb Bug fix Mar 18, 2019
seg_103_patch2image.ipynb commenting Apr 8, 2019
test_101_image_based.ipynb commenting Apr 8, 2019
test_102_patch_based.ipynb Bug fix Mar 18, 2019

README.md

Patch-to-Image Fully Convolutional Networks training for Retinal image segmentation

This repository contains the IPython code for the paper

Taibou Birgui Sekou, Moncef Hidane, Julien Olivier and Hubert Cardot. From Patch to Image Segmentation using Fully Convolutional Networks - Application to Retinal Images.

Given a retinal image database and a fully convolutional network (FCN) f, this tool first pre-trains it on an on-the-fly generated patch database, then fine-tunes it on the original full-sized images.

Framework

Setup

Environment: The following software/libraries are needed:

Datasets: The following datasets are used in our experiments:

Data preprocessing: All the images are preprocessed using:

  • Gray scale conversion
  • Gamma correction (with gamma=1.7)
  • CLAHE normalization
You can’t perform that action at this time.