Events
Building Capacity and Collaborations for Healthcare AI through Datathons | Institute of Artificial Intelligence for Digital Health
The research problems of health care extend beyond clinical medicine and cannot be solved by physicians working in isolation. Clinicians are well aware of the uncertainties and information gaps that permeate the practice of medicine, which range from the trivial but annoying to the significant and seemingly intractable. However, busy clinicians typically do not have the time, energy, or training to address questions they encounter in day-to-day practice. Instead, data scientists, Al scientists and engineers carry out health care Al research designed to decipher Clinical problems, but these efforts often are isolated from critical clinical input. Thus, over the years, together with the MIT Critical Data, my team at NUS have developed datathon as a platform in which diverse professionals (entrepreneurs, data scientists, Al engineers) work together on real clinical challenges and with real-world clinical data over a short period of time. We have so far organized 48 datathons around the globe We noticed that datathons can be an effective format to develop cross- disciplinary collaborations that lead to successful Healthcare Al projects. All these datathons helped my lab to develop productive research in the areas of medical imaging, reinforcement learning for treatment recommendation and language models for virtual clinical assistance.
Speaker Bio
Dr. Mengling Feng is currently a faculty member at Saw Swee Hock School of Public Health and the assistant director of research at Institute for Data Science at NUS. He is also the senior assistant director of National University Hospital, championing the big data analytics and healthcare Al initiatives. His research is to develop machine learning algorithms to extract actionable knowledge from large amount of data to enable better quality of healthcare. His research brings together concepts and tools across deep learning, optimization, signal processing, statistical causal inference and big data management. Dr. Feng’s work was recognized by both well-established journals, such as Science Translational Medicine, JAMA and top international conferences, such as KDD, AAAl and AMIA.