Dr. Ramakanth Kavuluru is a professor of biomedical informatics (Department of Internal Medicine) in the College of Medicine at the University of Kentucky (UKY). He also has a joint courtesy appointment in the Department of Computer Science at UKY. He graduated with a PhD in computer science in 2009 from UKY with a focus on the security properties of pseudorandom sequences. Subsequently, he worked in knowledge-based search systems for focused bioscience domains as a postdoctoral scholar at Wright State University. Since 2011, he has been working as a faculty member at UKY focusing on natural language processing methods and their use in biomedicine and healthcare.
High-level applications of Dr. Kavuluru’s research include cohort selection for clinical trials, literature-based knowledge discovery, computer-assisted coding, social media-based surveillance for substance abuse, and clinical-decision support for precision medicine. He employs methods from machine learning (including deep learning) and data mining fields to drive his research agenda. His recent methodological contributions deal with zero-shot and few-shot classification, large language models, transfer learning, domain adaptation, and end-to-end relation extraction. Thus far, in his capacity as primary advisor, he has helped seven doctoral students and 10 master’s students attain their graduation.
Predicting disease onset ahead of time is an important application of artificial intelligence (AI) and this is being actively pursued in the U.S. and other western nations. From a global health perspective, it is not clear if the implications of the findings of U.S. patient-based modeling translate to more populous and diverse areas of the world. Thus, using latest machine learning methods and data sets from Indian healthcare facilities, Dr. Kavuluru’s Fulbright-Nehru project is rigorously assessing how well the promise of AI holds when applied to the Indian patient setting compared to the simpler standard-of-care approaches to risk stratification.