The research group of Professor Kazuya Fujiwara and Professor Hirohito Sone of the Graduate School of Medical and Dental Sciences, Niigata University is an artificial intelligence that can make policy decisions at the specialist level beyond non-specialized general doctors in the clinical setting of life-threatening diabetes. (AI) Developed a system.It is realized by machine learning large-scale data of treatments performed by specialists in the past.
According to Niigata University, the research group has made AI machine-learn the medical record big data held by the Diabetes Data Management Study Group, and from 2009 to 2015, it will be used for the initial treatment of 20 people aged 4,860 and over who chose the drug by a diabetes specialist. They were asked to decide if insulin therapy was needed.
As a result, the correct answer rate for general practitioners was 43%, while AI gave 86%, which is almost double the correct answer.The research group believes that it is possible to judge whether insulin is necessary for initial treatment at a level close to that of a specialist, and showed that AI may be useful as a support role when general practitioners have to make policy decisions independently. It is supposed to be.
Lifestyle-related diseases such as diabetes continue to increase, making it difficult for specialists to treat all patients.Especially for diabetes, oral medication is often used as the first drug therapy, but depending on the condition, insulin injection at the first visit may be necessary.Misjudgment can lead to serious life-threatening consequences such as coma, which can lead to difficult decisions for general practitioners.