This paper proposes a context-aware deep learning approach for brain tumor segmentation, subtype classification, and survival prediction using multimodal magnetic resonance images (mMRI). A 3D context-aware deep learning model addresses the uncertainty of tumor location, followed by a 3D convolutional neural network (CNN) for subtype classification and a hybrid deep learning/machine learning method for survival prediction. Evaluated on the BraTS 2019 and CPM-RadPath 2019 datasets, the method shows robust performance, achieving second place in the CPM-RadPath 2019 challenge's testing phase.
Publisher
Scientific Reports
Published On
Nov 12, 2020
Authors
Linmin Pei, Lasitha Vidyaratne, Md Monibor Rahman, Khan M. Iftekharuddin
Tags
brain tumor
segmentation
subtype classification
survival prediction
deep learning
magnetic resonance images
Related Publications
Explore these studies to deepen your understanding of the subject.
