Title: Deep Learning in Brain Quantification and Cancer Radiotherapy
Speaker : Prof. Dinggang Shen (University of North Caroline at Chapel Hill)
Date: May 2 (Thurs.), 2019
Time: 11:00 AM
Location: Room 604, Woo Jung Information & Communication bldg.
Hosted by Dept. of Brain & Cognitive Engineering, Korea University / Center for Artificial Intelligence Research, Korea University / Institute of Brain and Cognitive Engineering, Korea University / BK21+ Global Leader Development Division in Brain Engineering, Korea University / Interdisciplinary Major in Brain and Cognitive Science
This talk will introduce our recent deep learning work on brain quantification and prostate cancer radiotherapy. Specifically, for automatic quantification of early brain development in the first year of life, i.e., with the goal of early identification of brain diseases such as autism, deep learning based brain image segmentation and cortical surface parcellation have been developed. For early diagnosis of Alzheimer’s Disease (AD) with the goal of possible early treatment, deep learning has been applied to unsupervised brain registration for precise inter-subject comparison and distinctive-regions based disease diagnosis. Besides, for effective treatment of prostate cancer, especially for MRI-based cancer treatment, a novel context-aware GAN (Generative Adversarial Networks) has been developed for synthesizing CT from MRI. Also, two novel deep learning techniques have been developed for automatic and precise segmentation of pelvic organs from the planning CT images to better guide radiotherapy. Both the clinical significance of each medical problem and the motivation of each developed technique will be clarified in this talk.