Diabetes AI adjustable for surveilling other public health concerns

Researchers have used machine learning to track diabetes at the population level.

While only modestly accurate, their technique may be generic enough to customize for keeping an eye on other widespread health conditions.

The work was conducted in France and is described in a study published Sept. 22 in Archives of Public Health.

Lead author Romana Haneef, PhD, of Santé Publique France, the French national public health agency, and colleagues trained an algorithm on data from more than 44,650 persons with diabetes in a national database.

They identified more than 3,400 variables and used 23 of the most diabetes-specific to refine the training of various algorithms.

The best-performing of these was a linear discriminant analysis model that crunched reimbursement data with an eye out for such variables as biological tests, drugs, medical services and hospitalizations.

The algorithm was 67% accurate at estimating diabetes prevalence, with sensitivity of 62% and specificity of 67%.

While this performance represents only a moderate success level, the authors suggest the work has “highlighted important methodological steps to apply machine learning techniques and their implications on large health administrative databases.”

“The next step is to apply this algorithm on the French National Health Database to estimate the incidence of type 2 diabetes cases,” Haneef and co-authors write. “More research is needed to apply various machine learning techniques to estimate the incidence of various health conditions and to calculate the contribution of various risk factors on developing type 2 diabetes.”

The study is available in full for free.

Dave Pearson

Dave P. has worked in journalism, marketing and public relations for more than 30 years, frequently concentrating on hospitals, healthcare technology and Catholic communications. He has also specialized in fundraising communications, ghostwriting for CEOs of local, national and global charities, nonprofits and foundations.

Trimed Popup
Trimed Popup