Hierarchical brain tumour segmentation using extremely randomized trees

作者：

Highlights：

• Hierarchy of Extremely Randomized Forest for glioma segmentation.

• Appearance and context features, some extracted over non-linearities applied to MRI.

• Low Grade Glioma data augmentation, for a balanced training set.

• Top Dice score for core and enhanced regions in BRATS 2013 Challenge and Leaderboard.

摘要

•Hierarchy of Extremely Randomized Forest for glioma segmentation.•Appearance and context features, some extracted over non-linearities applied to MRI.•Low Grade Glioma data augmentation, for a balanced training set.•Top Dice score for core and enhanced regions in BRATS 2013 Challenge and Leaderboard.

论文关键词：Brain tumour,Magnetic resonance imaging,Image segmentation,Hierarchy of classifiers,Extremely randomized trees,Machine learning

论文评审过程：Received 12 May 2017, Revised 28 March 2018, Accepted 5 May 2018, Available online 7 May 2018, Version of Record 15 June 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.05.006