AI Medical Model Gives More Cost-Effective and Accurate Results Than Traditional Treatment Methods

By: Zain Nabi  |   February 12th, 2013   |   Health, Living, News

Imagine you go to a computer instead of a doctor when you suffer from a disease. Rather than telling your feeling and condition to the doctor you feed the data in the computer, which uses its artificial intelligence to process the data and identify your illness and then come up with a solution to treat you. If it happens and people start trusting machines for their treatment, it would reduce the healthcare expenditure significantly.

 

Well, the good news is that many steps are being taken to make the above stated hypothesis a reality, and researchers from Indiana University have come quite close to achieving this objective. Casey Bennett and Kris Hauser have designed a model using an artificial intelligence framework to replace computers with doctors. Their research shows that the artificial intelligence framework can reduce the healthcare costs by over 50 per cent and improve patient outcomes by nearly 50 per cent.

 

Combining Markov Decision Processes and Dynamic Decision Networks, the framework is able to take decisions based on the data available and in case of availability of any information previously unavailable the framework adjusts itself and gives results accordingly. Bennett says:

 

“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.”

 

This framework is designed not for any particular disease. In fact the information fed in it can diagnose any disease based on the symptoms and data provided. Bennett is confident that this new approach is better than the traditional methods of treatment. He says:

 

“We’re using modern computational approaches to learn from clinical data and develop complex plans through the simulation of numerous, alternative sequential decision paths. The framework here easily out-performs the current treatment-as-usual, case-rate/fee-for-service models of health care.”

 

The two researchers had access to clinical data of over 6,700 patients. They randomly selected 500 patients from that group and fed their data into the system they had designed. Later, the results were compared with the ones that the doctors had come up with. One of the highlighting aspects of this development is the difference between the costs of two procedures. While the traditional method charged around $497 per unit, the artificial intelligence framework’ cost was $189.

 

Photo: IU