AI could help predict future diabetes cases
Image credit: Kiatiopl Kumphoo | Dreamstime
Scientists have developed a new technology that uses artificial intelligence (AI) to predict whether a patient will develop diabetes.
The preliminary research, which is yet to be published, applies a machine learning method to assess the risk of a person developing the lifelong condition that causes high blood sugar levels.
According to charity Diabetes UK, around 3.9m people in the UK are living with diabetes. The condition is linked to increased risks of other severe health problems, including heart disease and cancer. Experts have said that preventing diabetes to begin with is essential to reduce the risk.
“Currently, we do not have sufficient methods for predicting which generally healthy individuals will develop diabetes,” said Dr Akihiro Nomura, of the Kanazawa University Graduate School of Medical Sciences in Japan.
As part of the study, the researchers at the Japanese graduate school investigated whether machine learning could be used to diagnose diabetes.
Nomura and his colleagues analysed 509,153 nationwide annual health check-up records from 139,225 participants from 2008 to 2018 in the city of Kanazawa. 74,000 diabetes patients were among those included.
The data collected from medical records included physical exams, blood and urine tests and participant questionnaires. Patients without diabetes at the beginning of the study who underwent more than two annual health check-ups during this period were included. The team also made note of new cases of diabetes recorded during the patients’ annual health checks.
Dr Nomura and his colleagues then used the data to train a machine learning algorithm to predict those at risk of developing diabetes in the future.
As a result, the researchers identified a total of 4,696 new diabetes patients (7.2 per cent) in the study period. Furthermore, their trained computer model predicted the future incidence of diabetes with an overall accuracy of 94.9 per cent.
“Using machine learning, it could be possible to precisely identify high-risk groups of future diabetes patients better than using existing risk scores,” Nomura said. “In addition, the rate of visits to healthcare providers might be improved to prevent the future onset of diabetes.”
Nomura says he next plans to perform clinical trials to assess the effectiveness of using statins to treat groups of patients identified by the machine learning model as being at high risk of developing diabetes.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.
