Back to top

Requirements and Limits of Explainable AI in Healthcare

Monday, November 11, 2019 - 12:00

CTSI-CN Informatics Seminar Series with Muhammad Aurangzeb Ahmad 

Monday, November 11, 2019
12:00 – 1:00 pm

GWU, Office of CEHP – Watergate Office Building 
1st floor Conference Room 2600
Virginia Ave. NW, Washington, DC 20052

Lunch will be provided. 

Requirements and Limits of Explainable AI in Healthcare

Muhammad Aurangzeb Ahmad 
Associate Professor, University of Washington 
Principal Research Scientist, KenSci Inc. 

As AI and Machine Learning are being increasingly integrated into healthcare, challenges in creating responsible AI systems that are interpretable, fair, transparent, unbiased, robust and reliable are coming under tighter scrutiny. This talk will focus on one important aspect of such systems – Explainability – and will explore what constitutes explainable AI in healthcare, and what are the nuances, challenges, and requirements for its design. Drawing on insights from both academia and industry, the talk will delve into why explainability in healthcare is different from other domains and how deployed systems may have counterintuitive implications: e.g., explainability leading to distrust, explainable systems leading to more opaque models, etc. We will discuss the application of explainability techniques and the practical challenges in creating effective explainable AI models in healthcare. Finally, open problems and research directions for AI in the healthcare community will be described.


If you have any questions, please contact Liz Workman