Article by Barbara Toman
A machine can never replace your doctor. But artificial intelligence—the type of technology behind smart speakers and ride-hailing apps—is now being used to improve medical care.
Mayo Clinic is helping to lead that effort through a National Science Foundation (NSF) grant aimed at accelerating the application of artificial intelligence, also known as augmented human intelligence, to patient care.
The Mayo Clinic Center for Individualized Medicine (CIM) is co-sponsoring a workshop on July 19 for clinicians, academicians and industry groups to discuss opportunities in this transformative field. "Artificial Intelligence in Medicine—The Future Is Now" will focus on the real-world successes and challenges of bringing artificial intelligence into clinical practice.
"Our goal is to spark additional collaboration and the sharing of ideas that will move the technology forward," says Liewei Wang, M.D., Ph.D., co-principal investigator for the NSF grant. "Physicians and industry groups bring different perspectives. Working together is the way for us to advance applications of the science."
The workshop's co-sponsor, and Mayo Clinic's partner in the NSF grant, is the University of Illinois at Urbana-Champaign, a renowned center for computer science and engineering. Mayo Clinic and the University of Illinois together have established the Center for Computational Biotechnology and Genomic Medicine to engineer and optimize computing systems needed by industry and medicine for genomic analysis.
Artificial intelligence scans data and uses statistical methods, probability theory, and machine and deep learning to find patterns that are difficult for the human mind to see. This high-level computing augments physicians' knowledge to help doctors make predictions and treatment recommendations that are personalized for individual patients.
The workshop will include interactive panel and discussion sessions, and a keynote presentation by Andrew H. Limper, M.D., associate dean of Mayo Clinic's Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery. The workshop concludes at 4 p.m., coinciding with the start of a Medical Alley Chat networking event in Rochester.
Successes and challenges
"Artificial Intelligence in Medicine" will showcase examples of the successful use of artificial intelligence in such fields as neurology, radiology, oncology, psychiatry and cardiology.
One potential application involves the use of machine learning methods, developed by Mayo Clinic and Illinois, to locate the origin of seizure activity in the brains of people with epilepsy. Identifying that location makes successful treatment likelier for people whose epilepsy doesn't respond to medication.
The location and timing of seizure onset vary considerably among individuals. Monitoring the brain to localize and predict seizure onset has produced very promising results. Researchers have developed several biomarkers—measurements of the brain's non-seizure electrical activity—that can pinpoint seizure origin and forecast the occurrence. But it isn't clear which of these different biomarkers works best for a particular individual.
The methods developed by Mayo Clinic and Illinois use machine learning to combine complementary information from multiple biomarkers. In a recently published exploratory study, the researchers demonstrated improved localization and forecasting of seizures.
"Our study provides the first large-scale evaluation to our knowledge of an artificial intelligence-based approach for locating seizure origins from non-seizure data," says Yogatheesan Varatharajah, a Ph.D. candidate at Illinois and the study's first author. "When used in concert with multiple biomarkers, the method can outperform single biomarker-based approaches."
“In addition, our team has recently demonstrated the application of artificial intelligence to help determine when, where and how to stimulate the brain to improve memory function," adds Gregory Worrell, M.D., Ph.D., co-director of Mayo's Epilepsy and Neurophysiology Laboratory.
Other potential applications for artificial intelligence include:
"Much of this work is at a research stage. But once we develop these tools, we want the ability to apply them clinically to help patients," Dr. Wang says. "That's why it is so important to discuss the future of artificial intelligence and how industry can be involved."
One of the major challenges is building a workforce that can apply artificial intelligence to medicine. "With Illinois, Mayo Clinic is thinking about how we can systematically attract more students and junior faculty into applying their artificial intelligence expertise to medicine," Dr. Wang says. "These experts are in extremely high demand. We need them to work hand-in-hand with clinicians and biologists to keep moving the technology forward."
Translating technology and innovation into care that benefits patients is one of Mayo Clinic's core values. CIM is an outgrowth of that value, created to integrate the latest in genomics technology into clinical practice. Artificial intelligence is a key tool in that effort.
"Artificial intelligence is a catalyst and incubator for bringing individualized medicine technologies into our practice," Dr. Wang says. "We need to embrace artificial intelligence—but also use it wisely. That means identifying areas of medicine that can benefit from it the most."
Join the conversation
For more information on the Mayo Clinic Center for Individualized Medicine, visit our blog, Facebook, LinkedIn or Twitter at @MayoClinicCIM.
Learn more about the latest clinical applications of precision medicine at this year’s Individualizing Medicine Conference. It will be held Sept. 12-13, 2018.
Tags: #Artificial Intelligence, #Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Augmented human intelligence, Dr. Andrew Limper, Dr. Gregory Worrell, Dr. Liewei Wang, Mayo Clinic Center for Individualized Medicine, Precision Medicine, Univerity of Illinois at Urbana Champaign