Convolutional Neural Networks For Automated Edema Segmentation in Patients With Intracerebral Hemorrhage

Lucas A. Ramos, Dyantha G. van der Sluijs, Irem M. Baharoglu, Yvo B. Roos, Charles B. Majoie, Ludo F. Beenen, Renan S. de Barros, Silvia D. Olabarriaga, Henk A. Marquering

Apr 11, 2018 (modified: May 16, 2018) MIDL 2018 Abstract Submission readers: everyone
  • Abstract: Intracerebral hemorrhage (ICH) is a common type of stroke with high morbidity and mortality rate. Edema often forms around ICH. Because edema increases the chance of poor outcome, edema quantification is needed for finding the optimal ICH treatment. CNN has been proven to be a reliable method in medical image segmentation. In this study, we introduce CNN to develop an automated method for edema and ICH quantification. We found that our CNN is a promising quantification method for edema.
  • Author affiliation: Academic Medical Center
  • Keywords: Brain edema, convolutional neural networks, intracerebral hemorrhage
0 Replies

Loading