�ٺ���Ƶ

NYU Langone Health hosted its first annual Datathon and second annual Health Tech Symposium at its Science Building, located at 435 East 30th Street, in New York City. On May 9, attendees of the Health Tech Symposium listened to the keynote speeches on the main stage and participated in a variety of hands-on experiences, including a virtual reality, mentoring at an artificial intelligence (AI) and Machine Learning Data Bar, as well as three lectures from leaders in the healthcare technology space.

On May 10 and 11, a group of clinicians, data scientists, and engineers worked with mentors from Massachusetts Institute of Technology and Google to find new ways to answer actual clinical questions using large datasets from electronic health records.

“Today we are joined by a community of nearly 800 health technology, artificial intelligence, data science, and design enthusiasts—all gathered to share ideas, discuss trends, and learn new skills,” said Nader Mherabi, senior vice president and vice dean, chief information officer at �ٺ���Ƶ Health, in his introductory remarks. “At �ٺ���Ƶ Health, our goal is to transform the digital patient and clinician experience. We want to offer digital tools to our communities that meet their specific needs, that make their lives easier, and that feel more personal and considerate. We want to empower our patients, their families, and our team of care givers. That means leveraging these new techniques, and doing so in a way that puts the priorities of our patients, their families, and our clinicians first.”

Health Tech Symposium

Phillip Nelson, director of engineering of Google Research, was the keynote speaker and discussed his vision for the ways that machine learning can be used to advance healthcare and lessons learned from cases where machine learning was improperly applied. Yvonne W. Lui, MD, director of AI in �ٺ���Ƶ’s and project lead of �ٺ���Ƶ’s collaboration with Facebook’s Artificial Intelligence Research group discussed the latest applications of AI in imaging. Together, after their individual talks, audience members were able to participate in a fireside chat between the two speakers.

“We’re not here just to replicate or improve on human capabilities, but maybe even create entirely new capabilities,” said Dr. Lui. “AI is going to change the way we practice medicine. There is a huge need for partnership and cooperation among data scientists, biologists, medical experts, experts in law and bioethics, and so on.”

Following the keynote, 8 teams of aspiring entrepreneurs made their pitches for a $30,000 prize in the Biomed Venture Challenge, sponsored by . Each team posed clinical questions to the audience, explaining their process and what they hope to learn. The app team won the challenge for its digital platform that transforms the way that parents support young children’s mental health by giving parents the tools to understand, monitor, and support their young child’s emotional and behavioral challenges in their homes. The team is comprised of Helen L. Egger, MD, the Arnold Simon Professor of Child and Adolescent Psychiatry, chair of the , and director of the Child Study Center; Tim L. Verduin, PhD, clinical assistant professor in the Department of Child and Adolescent Psychiatry, director of Technology and Innovation, and clinical director of the Attention Deficit Hyperactivity Disorder (ADHD) and Behavior Disorder Service; as well as additional support team members.

Off the main stage, additional sessions, lectures, and hands-on experiences took place throughout the day. Sessions included the following:

• Experience VR: The Inevitability of Immersive Computing, Greg Dorsainville, �ٺ���Ƶ Health
• AI and Machine Learning Data Bar, , and Eduardo Iturrate, MD, �ٺ���Ƶ Health
• Data Science for the Non-Data Scientist, , �ٺ���Ƶ Health
• How to Publish a Clinical Machine Learning Paper: Relevant Questions at the Intersection of Medicine and Data Science, Dr. Aphinyanaphongs
• Data: The Rocket Fuel for Artificial Intelligence in Medical Research, Thomas Henson, Dell Technologies
• The Many Uses of Data at �ٺ���Ƶ Health, Dr. Iturrate

Datathon

During a 24-hour period, Datathon participants worked in teams to find new ways to answer actual clinical questions using large datasets from electronic health records. Leveraging the MIMIC Critical Care Database, the interdisciplinary teams collaborated to answer questions developed by the teams using data science techniques. MIMIC is an openly available dataset developed by the MIT Lab for Computational Physiology, comprising de-identified health data associated with approximately 40,000 critical care patients.

Following their marathon working session, teams presented their findings to a panel of judges, and one team was named the winner of the first annual competition.

This healthcare Datathon was the first ever held in New York City.

Media Inquiries

Kate Malenczak
Phone: 646-754-7367
kate.malenczak@nyulangone.org