Published May 3, 2021 | Version v1
Software documentation Open

MRI bias field correction with an implicitly trained CNN - Trained model and minimal example

  • 1. Umeå Universitet

Description

A software documentation for the paper "MRI bias field correction with an implicitly trained CNN".

The code is available at: https://github.com/attilasimko/bfc

The primary purpose of the repository is to publish the trained model, which is contained in the "implicit.zip" file  as a standalone model called "implicit.pb".

The repository also contains "MICE_example.ice" which is an example in MICE Toolkit where the implicit.pb model is compared to N4ITK using the dataset contained in data.zip, which was generated from BrainWeb (https://brainweb.bic.mni.mcgill.ca/).

Files

data.zip

Files (31.9 MB)

Name Size Download all
md5:e377161da524940f2061e3bf337cebc9
30.7 MB Preview Download
md5:fde06fda2dc14e15407611d50971188e
1.2 MB Preview Download
md5:983124939162c49a5e5d44853430def7
14.1 kB Download