Mike Livermore speaks with ICA4 Fellow Suranga Kasthurirathne, an Assistant Professor of Pediatrics at the Indiana University School of Medicine. Kasthurirathne is also a Research Scientist at the Clem McDonald Center for Biomedical Informatics at Indiana University’s Regenstrief Institute, and his work focuses on data analytics and machine learning in the healthcare context.

To begin with, Kasthurirathne offers some insight into his background, and describes how his experience developing health information infrastructure for under-resourced nations inspired an interest in data analytics. This leads to a discussion about the divide that exists between wealthy nations and poor nations, the economic disparities that exist within wealthy societies, and how artificial intelligence and information technology can help to bridge those gaps (:55 – 8:56). Kasthurirathne then speaks about how he uses AI in his research analyzing unstructured text data and natural speech, and the various tools that are used to analyze this data. One of the challenges that healthcare data analysts face is the strict rules regarding protection of medical records, meaning that there are significant barriers to access to data (9:00 – 26:17). Kasthurirathne goes on to explain the approaches that are taken to establish the models that data analysts use in their daily work. This part of the discussion also talks about the types of modeling systems healthcare data analysts use, and the positives and negatives of different systems. Kasthurirathne also explains the role predictive models play in the function of healthcare. (26:38 – 42:48) He also goes on to discuss the “garbage in, garbage out” problem of bias in models, the steps scientists are taking to avoid negative bias, and explains how some kinds of “bias” in modeling systems is not simply tolerable but, in fact, beneficial. This leads to a projection of how AI will evolve in the years to come, including how AI and data analytics may be able to evaluate the efficacy of different treatments and procedures, something that has already begun in some places (42:50 – 52:17). The conversation wraps up with Kasthurirathne describing the challenges his industry has faced in the wake of the Covid crisis, and how healthcare data analysts have responded to those challenges (52:43 – 1:00:23).

The Intercontinental Academia (ICA) is a global network of future research leaders sponsored by the University-Based Institutes of Advanced Studies. The ICA4 explores the complementarities between artificial intelligence and neuro/cognitive-science.