Using artificial intelligence and simulation modeling, researchers from Indiana University have found that machine learning can improve cost and quality of healthcare in the U.S., according to an announcement from IU.
The artificial intelligence framework, which combines Markov Decision Processes and Dynamic Decision Networks, used by IU researchers, shows how simulation modeling that "understands and predicts" the outcomes of health treatments could reduce healthcare costs and improve patient outcomes by about 50 percent. Their work has been published in the journal Artificial Intelligence in Medicine.
The research expounds previous work by the study's authors, which showed how machine learning could decide a patient's best treatment. The approach is not disease-specific and could work for any diagnosis or disorder.
"The Markov Decision Processes and Dynamic Decision Networks enable the system to deliberate about the future, considering all the different possible sequences of actions and effects in advance, even in cases where we are unsure of the effects," researcher Casey Bennett said in the announcement.
The study's authors conclude that an artificial intelligence simulation framework "can approximate optimal decisions, even in complex and uncertain environments.
"Future work is described that outlines potential lines of research and integration of machine learning algorithms for personalized medicine," they said.
FierceHealthIT's recent interview with Tina Buop, CIO of La Clinica de la Raza in Oakland, Calif., highlighted the importance of predictive analytics in healthcare. "You can bankrupt an organization very quickly if you don't understand your patient population," Buop said. Predictive analytics are used to know where to accept risk and how best to help a patient in advance. For example, Buop said, you can't fix your balance sheet when getting paid a fixed amount for a patient per month when they're your highest utilization.
Buop will join a panel at HIMSS13 in New Orleans hosted by FierceHealthIT and the College of Healthcare Information Management Executives, that will focus on using predictive analytics to improve care and efficiencies, including how to use data, artificial intelligence and clinical decision support to identify and create customized interventions for patients who are most at risk for adverse events and readmissions.
A study published in the International Journal of Medical Informatics in January highlighted how predictive modeling could reduce unnecessary lab tests for intensive-care patients with gastrointestinal bleeding.
To learn more:
- read the announcement from IU
- here's the the study's abstract
Related Articles:
How predictive analytics data helps your patients--and your business
Predictive analytics can slash unnecessary GI lab tests by half
Computational models increasingly improving research, healthcare quality