The Deep Learning Behind Radiology (IBM and Zebra Medical)

We will have two lectures (approx. 35-40 minutes each) in Hebrew. Light refreshments will be served before the first lecture.

Estimated schedule:
1800 refreshments at the J&J conference hall (3rd floor)
1830 First lecture
1920 Second lecture

First lecture:

Lecturer: Dr. Ayelet Akselrod-Ballin, Medical Imaging Research & Technology Lead at IBM Research.

Title: Will AI change Radiology?

Artificial intelligence is rapidly evolving in the healthcare domain. Recent advances in deep learning have reported notable results including pneumonia detection on chest X-Rays, classification of skin cancer, and detection of diabetic retinopathy in retinal fundus photographs. Medical Sieve, is an ambitious long-term exploratory IBM grand challenge to build a next generation cognitive assistant with advanced multimodal analytics, clinical knowledge, reasoning and deep learning capabilities that is qualified to assist in clinical decision making in radiology. This talk will present some of the challenges and results obtained throughout this project, focusing on the breast imaging domain.

Dr. Ayelet Akselrod-Ballin research focuses on developing novel technologies for computer vision, machine learning, deep learning and biomedical image analysis. Ayelet did her postdoctoral studies in the Computational Radiology Laboratory at Harvard Medical School, Children's Hospital (Boston). She received her PhD in applied mathematics and computer science in 2007 from Weizmann Institute of Science, and her MSc in computer science (magna cum laude) and a BSc in physics and computer science (magna cum laude) from Tel-Aviv University, Israel. Prior to joining IBM she was the computer vision and algorithms team leader at the Israeli Ministry of Defense and an image analysis expert at Aposense Ltd.

Second lecture:

Lecturer: Dr. Jonathan Laserson, Lead AI researcher at Zebra Medical.

Title: Embrace The Noise: Mining Clinical Reports to Gain a Broad Understanding of Chest X-rays

The chest X-ray scan is by far the most commonly performed radiological examination for screening and diagnosis of many cardiac and pulmonary diseases. It also one of the hardest to interpret, with disagreement rating of around 30% even for experienced radiologists. At Zebra, we have access to millions of X-ray scans, as well as their accompanied anonymized textual reports written by hospital radiologists. Can this data be used to teach an algorithm to identify significant clinical findings from these scans? By manually tagging a relatively small set of sentences, we were able to construct a training set of almost 1M studies over the 40 most prevalent chest X-ray pathologies. A deep learning model was trained to predict the findings given the patient frontal and lateral scans. We compared the model's predictions to those made by a team of radiologists. Would the average radiologist agree more with his/her colleagues or with the model?

Dr. Jonathan Laserson is a senior researcher at Zebra Medical Vision and a Machine Learning expert and consultant. He has a PhD from the Computer Science AI lab at Stanford University and was a lecturer at Bar-Ilan University. After a few years doing machine learning at Google and IBM, today he is focused on Deep Learning algorithms and their application to medical images understanding.