Afternoon of Science Series: Department of Biomedical Informatics

The Afternoon of Science series at the Vagelos College of Physicians and Surgeons continued June 26 with presentations from the Department of Biomedical Informatics.

In her opening remarks, VP&S Dean Katrina Armstrong emphasized the importance of sharing research to inspire collaboration, noting that the events are increasingly attended by colleagues from across Columbia University.

“During these afternoon sessions, we’re not only celebrating the departments, but also weaving all this knowledge together, so that over the course of time we can come together to strengthen our collaborations, make the sum greater than the parts, and make major strategic investments in critical areas that are going to benefit science more broadly,” Armstrong said.

The event was hosted by Noémie Elhadad, chair of the Department of Biomedical Informatics, who spoke about the department’s history and the breadth of research. The department first began as a center in 1987, becoming the Department of Medical Informatics in 1994, and the Department of Biomedical Informatics in 2003. The highly collaborative department works with VP&S clinical departments, NewYork-Presbyterian, Columbia’s business school, and Columbia’s engineering school, among others, to produce theoretical and practical advances in AI, data science, and human-computer interaction, which will lead to better health and new biomedical knowledge.

“What we’re trying to achieve is a health ecosystem where humans and intelligence systems can interact with each other with aligned goals and values,” Elhadad said. “We hope that by building this type of technology, we can improve patient experience, health equity, and clinician well-being while also creating new biomedical knowledge. These different dimensions, we believe, will ultimately lead to better health for all.”

Attendees at the June 26 event included external advisers from Khoury College of Northeastern University, Stanford University, and CSAIL at the Massachusetts Institute of Technology.

Faculty presentations were made by:

• Pierre Elias, Assistant Professor of Medicine (in Biomedical Informatics)
  Presentation: “Develop, Validate, Deploy—The Practice of Cardiovascular AI”
  Elias’s lab develops machine learning technologies for medical imaging to improve the detection and management of cardiovascular disease. He is also a general cardiologist in the VP&S Division of Cardiology and medical director for artificial intelligence at NewYork-Presbyterian.
• Shalmali Joshi, Assistant Professor of Biomedical Informatics
  Presentation: “A Path to Safe, Reliable, and Equitable AI for Health Care”
  Joshi’s research is on the algorithmic safety of machine learning for human-centered domains. Joshi has contributed to explainability, robustness, and novel algorithms for machine learning safety, emphasizing practical generative settings and impact on decision-making.
• Gamze Gürsoy, Assistant Professor of Biomedical Informatics
  Presentation: “Privacy in the Era of Precision Medicine”
  Gürsoy’s research group develops privacy-preserving tools to analyze and understand large-scale omics data in relation to diseases and phenotypes, with a particular interest in developing software, file formats, and pipelines that enable broad sharing and analysis of sensitive genotypic and phenotypic data in public servers.
• Lena Mamykina, Associate Professor of Biomedical Informatics
  Presentation: “Making AI Useful: Human-Centered Approach to AI in Health and Medicine”
  Mamykina’s research interests include an individual’s sense-making and problem-solving in context of health management, collective sense-making within online health support communities, clinical reasoning and decision-making, communication and coordination of work in clinical teams, and ways to support these practices with informatics interventions. She also focuses on analysis of health information technologies and how they are used among critical care teams, as well as social computing platforms for facilitating knowledge sharing within clinical communities and within online health support groups.
• George Hripcsak, Vivian Beaumont Allen Professor of Biomedical Informatics
  Presentation: “Changing Health Care with Observational Research”
  Hripcsak’s research focuses on clinical information stored in electronic health records and the development of next-generation health record systems. Health record data are sparse, irregularly sampled, complex, and biased. Using nonlinear time series analysis methods borrowed from statistical physics, machine learning, knowledge engineering, and natural language processing, he is developing the methods necessary to support clinical research and patient safety initiatives using health record data. Hripcsak also has a long track record of developing, implementing, and studying informatics interventions to improve health care.