Communications of the ACM
Scientists Enlist Supercomputers, Machine Learning to Automatically Identify Brain Tumors
BRATS image shows tumor sub-regions: edema (yellow); non-enhancing solid core (red); necrotic/cystic core (green); and enhanced core (blue).
Credit: IEEE Transactions on Medical Imaging
Researchers at the University of Texas at Austin are leading a multi-institutional team in developing a fully automatic method that combines biophysical models of tumor growth with machine-learning algorithms to analyze magnetic resonance imaging data of glioma patients. All of the components of the new method were enabled by supercomputers at the Texas Advanced Computing Center.
The researchers tested their method in the Multimodal Brain Tumor Segmentation Challenge 2017, an annual competition in which research groups from around the world present methods and results for computer-aided identification and classification of brain tumors and other types of cancer.
The team's system scored in the top 25 percent in the challenge and was near the top for the whole tumor segmentation. "Our goal is to take an image and delineate it automatically and identify different types of abnormal tissue — edema, enhancing tumor, and necrotic tissue," says UT Austin professor George Biros.
From Texas Advanced Computing Center
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found