UT System symposium examines how AI can create ‘big solutions to big problems’
Inaugural event at UT Southwestern attracts top experts to discuss ethics, costs, and pace of AI innovations
Artificial intelligence is already being used in a multitude of ways in medicine – from routine computational tasks to deep machine learning and generative AI – all with the most human of pursuits in mind: to provide the best care for patients.
That was the overriding theme of the inaugural UT System AI Symposium in Health Care, held in late May on the UT Southwestern campus. More than 700 scientists, physicians, educators, students, and researchers representing all University of Texas health institutions and seven medical schools gathered in person and virtually to delve into the evolution of AI and how it could transform health care in the most effective, ethical ways.
“The question isn’t what part of medicine will AI affect, but rather, what part of medicine won’t be transformed by AI,” noted Eric Peterson, M.D., M.P.H., Professor of Internal Medicine, Vice Provost and Senior Associate Dean for Clinical Research at UT Southwestern, and one of the symposium organizers. “My sense is that just about every discipline of medicine will be changed, from how we do our research and make our discoveries to how we treat our patients to how we conduct our operations to, ultimately, how we train the next generation of medical care providers.”
The symposium, held May 30-31 in the Tom and Lula Gooch Auditorium on UTSW’s South Campus, included keynote addresses, panel discussions, poster sessions, and short plenary talks as well as a joint introduction from UT System Executive Vice Chancellor for Health Affairs John M. Zerwas, M.D., and UT Southwestern President Daniel K. Podolsky, M.D.
Both leaders talked about the potential of AI and the benefits it has already brought to medical care and clinic management. They also acknowledged the need for thoughtful guardrails for this emerging technology.
“With our esteemed faculty, talented students, cutting-edge facilities, and a strong spirit of collaboration, the University of Texas System is uniquely poised to make significant contributions in the field of artificial intelligence,” Dr. Zerwas said.
“This is a great time to think creatively about the ways in which AI can be used to transform our ability to provide care, to advance research, and to augment our approach to education and training,” added Dr. Podolsky. “We are also envisioning how these emerging technologies can be leveraged to improve operational efficiencies and enhance productivity.”
Greater efficiency, better care
During a Day 2 presentation titled Industry and Academic Partnership, one of the panelists looked across the stage at executives from tech giants Microsoft, Amazon, and Google, and commented: “Wow, the wattage up here is intense.” The same without a doubt could have been said for the symposium’s keynote speaker, Nigam Shah, M.B.B.S., Ph.D., Chief Data Scientist at Stanford Health Care, who talked about the challenges and opportunities of AI innovations.
“It’s easy to get caught up in the excitement of all that AI models can do,” said Dr. Shah, who has co-founded three companies and written more than 300 articles in scientific publications. “But we still have to ask the most important questions, such as, ‘Does it positively impact patients? Does it increase efficiency or decrease cost? Or both?’”
Dr. Shah emphasized that it was incumbent on the growing ecosystem of health care and technology organizations to make AI useful. “Whatever information AI provides,” he said, “people will still have to be able to have the capacity to act on it.”
Cost was another common theme in panel discussions and in questions from the audience.
“The huge potentials brought about by the AI revolution can significantly reduce the cognitive labor cost in health care by automating a myriad of routine cognitive tasks,” said Jiajie Zhang, Ph.D., Dean of the D. Bradley McWilliams School of Biomedical Informatics at UTHealth Houston, a panelist on the Day 1 AI Across the UT Health System discussion. “However, without changing the way we do the work and without a plan to retrain and upskill the workforce, we will not get the full benefit from AI.”
A day later, Alexander “Sasha” Sicular, Executive Lead of Healthcare and Life Sciences for Google Public Sector, asked members of the Industry and Academic Partnership panel: “Can we afford not to invest in AI, which is clearly an inevitable future, not just for health care but for every industry on this planet?” He shared an anecdote about how technology has evolved throughout his life, from the ’80s when he was watching movies on a VCR to beepers, cellphones, Wi-Fi, the cloud, and now AI.
“With every subsequent wave of technology, it comes with orders of magnitude of efficiencies, many of which we cannot even comprehend at this moment,” Mr. Sicular said. “So, again, I would ask many of the decision-makers in this room, ‘Can we afford not to invest in this area?’”
Ethical and equity issues
Responsibility, security, and the ethics of AI were themes that arose throughout the symposium, including during a panel discussion that featured Jonathan Efron, M.D., Executive Vice President for Health System Affairs at UT Southwestern.
Dr. Efron said that when choosing to implement AI in a health care setting, it’s critical to talk with patients and providers to target the most valuable areas to employ the technology “so that instead of picking up something four floors away, that nurse or doctor is now taking care of patients.”
He said he witnessed the power of AI when he worked at Johns Hopkins University School of Medicine, which had access to a system for analyzing methods of care that took 10 years to build. In that case, AI proved to be a literal lifesaver during the COVID-19 pandemic.
“This real-time data led to a clinical decision that people (with COVID-19) should be on their stomachs,” Dr. Efron said. “It was amazing – this had a specific impact on the decisions of how to care for people.”
Andrea Cooley, D.O., FACOS, a cardiothoracic surgeon at the UT Tyler School of Medicine, sounded notes of both caution and optimism about AI and the impact it could have on rural health care in her Equity by Design presentation.
“Digital solutions can’t help those who can’t access the internet,” said Dr. Cooley, who shared statistics about the lack of digital access in portions of East Texas. She also suggested that a unified data warehouse shared among UT institutions might be able to address serious health care deficiencies in that region, such as the lack of pediatric care – 27 counties in the area have no pediatricians at all, she said.
“A big system with big data can create big solutions to big problems,” Dr. Cooley said.