Casey Bennett and Kris Hauser, researchers from IU School Informatics and Computing demonstrated how artificial intelligence (AI) can predict patient treatment outcomes, determining the best treatment. Simulations using a combined AI showed how health care costs could be lowered by 50 percent and improve patient outcomes by 50 percent.
"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," said Bennett.
The computerized approach can help with the treatment of various ailments and can be used to diagnose a disorder by inputting relevant information. In the United States, healthcare faces many dilemmas and this study addresses them. By 2050, the rising cost of health care is expected to reach 30 percent of the United States' gross domestic product.
Patients are sometimes diagnosed incorrectly or receive the incorrect treatment, which can be prevented.
"We're using modern computational approaches to learn from clinical data and develop complex plans through the simulation of numerous, alternative sequential decision paths," Bennett said. "The framework here easily out-performs the current treatment-as-usual, case-rate/fee-for-service models of health care."
With access to patient clinical records, the researchers selected 500 out of 6,700 patients who had major health issues such as diabetes, hypertension and heart disease. Actual physician performance was paired with the AI's patient outcomes, using actual patient data. As a result, the AI's treatment cost per unit was $189 compared to the standard treatment of $497.
"This was at the same time that the AI approach obtained a 30 to 35 percent increase in patient outcomes," Bennett said. "And we determined that tweaking certain model parameters could enhance the outcome advantage to about 50 percent more improvement at about half the cost."
It was found that for complex treatment decisions, AI may be a better option than the traditional case by case approach.
