Executive Update
November 24, 2019

From Peter Embí, CEO: Regenstrief Driving AI, Leading to Transform Healthcare

Dr. Peter J. Embi

Artificial intelligence (AI) is not a new concept, but a recognition of its potential and the focus on AI is clearly growing across many industries. In healthcare, advances in AI are poised to transform care and the clinician-patient relationship.

A few early examples help set the stage for current ground-breaking work and let us look toward the future.

Vannevar Bush, who headed the U.S. Office of Scientific Research and Development during World War II, developed innovate ways to deal with huge quantities of information. Alan Turing conducted early work to formalize the concepts of algorithms and computation, and proposed in 1950 what we now call the Turing Test as a measure of machine intelligence. Numerous talented scientists created Deep Blue, the IBM computer that defeated world chess champion Garry Kasparov in 1997, and more recently Watson, the computer that triumphed on TV’s Jeopardy in 2011. These are just a few of the many contributors to the development of AI.

Over the past half-century, Regenstrief scientists have been pioneers in the use of AI and related methods — machine learning, natural language processing, computerized decision support and other innovative uses of big data — to improve healthcare delivery. Indeed, from its early days, the institute has been designing, developing, testing and applying original concepts and innovative approaches to enable computers to analyze large amounts of data, draw conclusions, and recommend actions based on the results.

Three years ago, Regenstrief’s Vice President of Data and Analytics (then CBMI director) Shaun Grannis, M.D., along with colleagues including Research Scientists Brian Dixon, PhD, Suranga Kasthurirathne, Ph.D. (then a grad student), Burke Mamlin, M.D., and former fellow Judy Gichoya, M.D., (now an assistant professor of radiology and of informatics at Emory University) concluded that machine learning had come of age in public health reporting. In a landmark study published in Journal of Biomedical Informatics, they found that existing algorithms and open source machine learning tools were as good as, or better than, human reviewers in detecting cancer cases using data from free-text pathology reports. The AI approach was also faster and less resource intensive than human efforts. While there remains much work to further develop, refine, validate and demonstrate widespread effectiveness of such approaches, and we are certainly not yet able to recommend that computers replace humans in most healthcare situations, it is increasingly clear that there are certain activities where computers will be able to outperform humans and do so more consistently. As Dr. Grannis said at the time, “We have come to the point in time that technology can handle this [analyzing free-text data]. A human’s time is better spent helping other humans by providing them with better clinical care.”

This summer at MedInfo, Drs. Grannis, Dixon and Kasthurirathne and Regenstrief Summer Scholar Gregory Dexter presented their latest work on machine learning approaches to public health focusing on employing AI to facilitate notifiable disease reporting.

Indeed, Regenstrief scientists are continually recognized as among the nation’s leaders in the study of application of AI to improve human health. Eneida Mendonca, M.D., PhD, our vice president for research development, who has devoted a large part of her career to machine learning, has participated in leading education and thought leadership conversations about AI, including last month, when she presented a keynote address on the state of AI research to attendees at a workshop in São Paulo hosted by Brazil’s National Institute of Science and Technology in Scientific Assisted Medicine.

Last week, Dr. Mendonca joined members of the National Academy of Medicine Leadership Consortium for a Value and Science-Driven Health System panel at the American Medical Informatics Association (AMIA) 2019 symposium discussing opportunities and challenges related to the expanded application of AI in healthcare. In early December, Dr. Mendonca, Dr. Umberto Tachinardi, Dr. Grannis and myself will be part of a panel on “algorithmovigilance” at AMIA’s Health Informatics Policy Forum. We will be exploring considerations for systematic monitoring and continuous learning in an AI-driven healthcare future.

In the coming months, Regenstrief will host a number of AI events designed to provoke conversations and solutions to move the field forward. The first event will be the Regenstrief Entrepreneurial Ecosystem Forum (REEF) “Putting the I in AI:  How healthcare innovators are using data to move from ideas to implementation” on December 3. You are invited to join us to hear from three companies who are accomplishing meaningful outcomes using data and AI. There is more information on signing up in the Regenstrief newsletter. This is just the first in an exciting series of discussions. More information to come on those events.

Following are just of few more exciting highlights of the current AI research landscape at Regenstrief.

Research Scientists Kun Huang, Ph.D., Zhi Han, Ph.D., and IUPUI colleagues have developed and used artificial intelligence tools to examine symptom clusters in cancer patients. Their seminal study won the Best Paper Award at the recent 10th Association for Computing Machinery (ACM) Conference on Bioinformatics, Computational Biology and Health Informatics. This prestigious prize is presented for a paper that represents groundbreaking research. Through the Best Paper Award, the ACM highlights theoretical and practical innovations likely to shape the future of computing.

Uppstroms, which has triumphed in several tech competitions including AMIA’s 2018 Pitch IT Competition, came out a winner again, this time at the American Health Information Management Association’s Health Data and Information Conference pitch competition this fall. The Uppstroms app uses AI — specifically machine learning analytics — to address upstream social risk to promote better outcomes for patients and better health systems. Uppstroms’ predictive modeling algorithm analyzes social determinants of health to help clinicians identify patients who might benefit from wraparound services, such as those provided by a social worker or dietician. These services can help avoid preventable utilization of the health system and can reduce costs and improve patient outcomes. Uppstroms is already in use at Eskenazi Health. The Uppstroms team, which includes Regenstrief Research Scientists Joshua Vest, PhD,  Shaun Grannis, M.D., Suranga Kasthurirathne, PhD, Nir Menachemi, PhD, and Paul Halverson, DrPh, dean of the IU Richard M. Fairbanks School of Public Health at IUPUI, hope to expand its use to other health systems.

The possibilities for AI application in healthcare are nearly endless. As innovators in big data analysis, decision support, natural language processing, machine learning and other subfields of AI for the past half century, Regenstrief research scientists will continue leading the field, identifying, testing, applying and driving new AI solutions to improve human health locally, nationally and around the world.

Related News

Regenstrief researcher awarded $1.9 million CDC grant

Award funds study to improve long-term outcomes of those living with congenital heart defects Jill Inderstrodt, PhD, MPH, received

Debra Litzelman, M.D., and Glenda Westmoreland, M.D.

Addressing the geriatric healthcare workforce shortage

Successful new training method one of first to use virtual standardized patients INDIANAPOLIS – The pandemic has highlighted the

Health data standards experts from 20 countries participated in 2024 LOINC® conference

Health data standards experts from 20 countries participated in 2024 LOINC® conference

Representatives from reference labs, healthcare providers, government agencies, insurance companies, software and device manufacturers and researchers from 20 countries

2024 OpenHIE Community Meeting November 12-16 in Sri Lanka

OpenHIE, a global initiative focused on supporting the use of data exchange standards and the implementation of health system