Application of Deep Learning Techniques for Medical Image Analysis
April 25, 2018 at 1:30 PM to 3:30 PM
|Location:||5345 Herzberg Laboratories|
Medical imaging, (e.g., computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), mammography, ultrasound, X-ray) has advanced at a rapid speed over last decades. Currently, the medical image interpretation is mostly performed by human experts, which is a tedious task and subject to high inter-operator variability. Deep learning is providing exciting solutions for medical image analysis problems. Recent advances in deep learning have helped to identify, classify, and quantify patterns in medical images. In this seminar, we introduce the principles and methods of deep learning concepts, particularly convolutional neural network (CNN). We show how CNN operate. I will describe about several interesting applications of deep learning for medical image analysis, including one of my recent work on segmenting myocardial scar (injured) tissue in the heart from 3D LGE MR images of patients with heart disease.
Fatemeh Zabihollahy is currently a PhD candidate at Carleton University. She obtained her MASc (2016) and BASc (2001) both in Biomedical Engineering from Carleton University, Canada and Shahid Beheshti University, Iran, respectively. She worked in medical devices industry as an R&D engineer for ten years. Her research interest is in the field of application of deep learning techniques for medical image analysis.