search

dhirajmaji7 / QuickNAT-keras

Implementation of QuickNAT for brain tumor segmentation using Keras library with tensorflow as backend. The QuickNAT is a fast and accurate segmentation model. The model has been implemented along with the Unpooling and a combined loss function of dice score and weighted cross-entropy as described in the paper.
MIT License
1 stars 0 forks source link

readme

QuickNAT-keras

Implementation of QuickNAT for brain tumor segmentation using Keras library with tensorflow as backend. The QuickNAT is a fast and accurate segmentation model. The model has been implemented along with the Unpooling and a combined loss function of dice score and weighted cross-entropy as described in the paper.

Cite this paper:

Abhijit Guha Roy, Sailesh Conjeti, Nassir Navab, Christian Wachinger. QuickNAT: A fully convolutional network for quick and accurate segmentation of neuroanatomy, NeuroImage, Volume 186, 2019, Pages 713-727, ISSN 1053-8119, https://doi.org/10.1016/j.neuroimage.2018.11.042