Virtual Seminar Series

August 28, 2025
Via Zoom

A town hall for introductions, sharing the goals of the RIC, and getting input on building the community!

September 4, 2025
Via Zoom

Title: Data Overwhelm: Monitoring Digital Health Informatics Data for Critical Events

Abstract: Large volumes of health data are collected every day with the hope that these measurements can capture changes in health dynamics and serve as an early warning system (e.g. for hospitalization). However, the sheer volume of data makes these events more difficult to detect. We can frame this problem as a new AI task called multi-stream outlier ranking. I’ll describe methods for this task that work with the imperfect statistical properties of real-world healthcare data and how these methods have been deployed as part of a human expert-in-the-loop AI system for public health respiratory-illnesses. In our most recent evaluations, experts detected health-related events 288x more efficiently than without this system. As a result, thousands of public health data issues have identified across the United States.

 I’ll share our best-practices and lessons learned in setting up this system, and cover new research efforts to reduce data overwhelm in the mental health x AI space, specifically for clinicians working on mood disorders like depression or bipolar disorder.   

Bio: Ananya Joshi is an Assistant Professor in Psychiatry & Behavioral Science and Computer Science. Her group's research develops computational and Generative AI methods and systems for diagnosing, treating, and monitoring mood disorders. She previously received her Ph.D. and M.S. in Computer Science from Carnegie Mellon University, supported by the NSF Graduate Research Fellowship Program, and her B.S.E. in Computer Science from Princeton University.

September 18, 2025

Title: The Role of LLMs in Reproductive Health: Insights from Myna Mahila Foundation

Abstract: This fireside chat with Myna Mahila Foundation will reflect on how large language models (LLMs) are being leveraged in community health settings globally. We will discuss opportunities, risks, and practical applications of AI-driven health communication through the lens of Myna Bolo, a WhatsApp chatbot that supports question-answering on sexual and reproductive health in an underserved community in Mumbai, India. By bringing together multi-disciplinary and multi-stakeholder perspectives—from academia, design, and implementation—we hope to foster rich and critical discussion, and co-create a path forward for the thoughtful use of LLMs on sensitive and culturally-embedded health topics.