Efficient Multi-Scale 3D CNN with Fully Connected CRF for Accurate Brain Lesion Segmentation

18 Mar 2016Konstantinos KamnitsasChristian LedigVirginia F. J. NewcombeJoanna P. SimpsonAndrew D. KaneDavid K. MenonDaniel RueckertBen Glocker

We propose a dual pathway, 11-layers deep, three-dimensional Convolutional Neural Network for the challenging task of brain lesion segmentation. The devised architecture is the result of an in-depth analysis of the limitations of current networks proposed for similar applications... (read more)

PDF Abstract

Evaluation results from the paper

Task Dataset Model Metric name Metric value Global rank Compare
Brain Tumor Segmentation BRATS-2015 3D CNN + CRF Dice Score 85% # 1
Lesion Segmentation ISLES-2015 3D CNN + CRF Dice Score 59% # 1