Speaker: Krzysztof Jerzy Geras, NYU School of Medicine
UCL Contact: Dominique Drai (Visitors from outside UCL please email in advance).
Date/Time: 18 Jul 18, 13:00 - 14:00
Venue: Roberts 106
Advances in deep learning for natural images have prompted a surge of interest in applying similar techniques to medical images. The majority of the initial attempts focused on replacing the input of a deep convolutional neural network with a medical image, which does not take into consideration the fundamental differences between these two types of images. Specifically, fine details are necessary for detection in medical images, unlike in natural images where coarse structures matter most. This difference makes it inadequate to use the existing network architectures developed for natural images, because they work on heavily downscaled images to reduce the memory requirements. This hides details necessary to make accurate predictions. Additionally, a single exam in medical imaging often comes with a set of views which must be fused in order to reach a correct conclusion. In our work, we propose to use a multi-view deep convolutional neural network that handles a set of high-resolution medical images. We evaluate it on large scale mammography-based breast cancer screening (BI-RADS prediction) using 886,000 images. We focus on investigating the impact of the training set size and image size on the prediction accuracy. Our results highlight that performance increases with the size of training set, and that the best performance can only be achieved using the original resolution. In the reader study, performed on a random subset of the test set, we confirmed the efficacy of our model, which achieved performance comparable to a committee of radiologists when presented with the same data.
Krzysztof Jerzy Geras
Krzysztof recently started as an assistant professor at the NYU School of Medicine. He previously worked as a postdoctoral researcher at the NYU Center for Data Science with Prof. Kyunghyun Cho, Prof. Linda Moy and Prof. Sungheon G. Kim. His main interests are unsupervised learning with neural networks, model compression, transfer learning and evaluation of machine learning models. During his PhD studies at the University of Edinburgh he was supervised by Dr. Charles Sutton. Prior to studying towards a PhD in Machine Learning, he got BSc and MSc degrees in Computer Science from the University of Warsaw. He worked on his MSc thesis as a visiting student at the University of Edinburgh under the supervision of Dr. Amos Storkey, his second supervisor was Prof. Andrzej Szałas. His thesis was on Machine Learning Markets. He also did industrial internships in Microsoft Research (Redmond, working with Rich Caruana and Abdel-rahman Mohamed), Amazon (Berlin, Ralf Herbrich's group), Microsoft (Bellevue) and J.P. Morgan (London)