Events Calendar
Adam Yala (MIT): Rethinking Cancer Risk and Personalized Screening with Deep Learning
Tuesday 25 February 2020, 12:00pm - 01:00pm
In this talk, we'll discuss deep learning approaches to tackle problems across the spectrum of cancer screening. Specifically, we'll look at how deep learning on routinely collected information available like ICD-10 codes at population scales can be used to identify populations for pancreatic cancer screening. Next, given access to imaging, we will discuss careful technical design can yield state-of-the-art risk models, using our work in breast cancer as a case study. Finally, we will close with new and ongoing work to devise general personalized screening policies using reinforcement learning to leverage any risk model to achieve population screening objectives. While focusing on cancer guidelines and screening, this talk will span problems in text, vision and reinforcement learning.
Location : Goitein Room
