About this Class

About the Seminar

Artificial intelligence and machine learning have the potential to greatly transform healthcare. Although these techniques have shown remarkable performance for many tasks including medical image analysis, we will share some of the challenges that we have faced in developing robust and trustworthy algorithms including a lack of repeatability, explainability, generalizability, and the potential for bias. Access to large, representative, diverse, and well curated datasets is vital to improving the performance of machine learning algorithms. Historically, concerns related to patient privacy, regulations, cost and logistical challenges have limited data-sharing. Approaches such as federated learning can improve the robustness of algorithms by providing a framework where the trained models have been exposed to multi-institutional datasets without the need for data-sharing. We will review examples of privacy preserving learning from multi-institutional datasets and discuss successes as well as directions for future research.

About the Speaker

Jayashree Kalpathy-Cramer is currently an Associate Professor of Radiology at Harvard Medical School, and a Co-Director of the QTIM lab and the Center for Machine Learning at the Martinos Center. She is the incoming chief of the new Division of Artificial Medical Intelligence in Ophthalmology at the University of Colorado (CU) School of Medicine. An electrical engineer by training, she worked in the semiconductor industry for several years. After returning to academia, she is now focused on the applications of machine learning and modeling in healthcare. Her research interests include medical image analysis, machine learning and artificial intelligence for applications in radiology, oncology, and ophthalmology. The work in her lab spans the spectrum from novel algorithm development to clinical deployment. She is passionate about the potential that these techniques have to improve access to healthcare in the US and worldwide. Dr. Kalpathy-Cramer has authored over 200 peer-reviewed publications and has written over a dozen book chapters.

About the Seminar Series

The seminar is open to the public and registration is required each month. Individuals who need interpreting services and/or other reasonable accommodations to participate in this event should contact Rachel Pisarski(link sends e-mail) at 301-670-4990. Requests should be made at least five days in advance of the event.