Innovating Challenges and Experiences in Emory Health AI Bias Datathon: Experience Report.

Summary: Artificial Intelligence (AI) holds incredible promise for healthcare, but it also carries a risk: if the data used to train AI is biased, the medical advice it gives will be too. In August 2023, Emory University tackled this head-on by hosting the "Health AI Bias Datathon." This event wasn't just for computer whizzes; it brought together students, researchers, and healthcare professionals from diverse backgrounds to form teams.

This paper reports on how the event was built from the ground up—from registration to the final presentations. It highlights the resilience and creativity of the participants as they worked with complex datasets to identify and fix biases. The organizers surveyed the attendees afterward to learn what worked and what didn't. The result? A roadmap for running future events that foster innovation and help ensure health AI serves everyone fairly.