Artificial intelligence may be able to identify patients at risk for undiagnosed celiac disease before the disease presents itself, a study by Maccabi KSM Research and Innovation Center and Predicta Med found.

The findings, published in Nature Portfolio’s Scientific Reports journal, suggested that by providing machine learning models with electronic medical records (EMRs), they could predict celiac disease up to four years before diagnosis.

Celiac disease – an autoimmune disease affecting one’s ability to digest gluten – affects an estimated 1% of adults and children worldwide, with many individuals suffering from symptoms for years, even more than a decade, before receiving a diagnosis.

In the study, which received ethical approval from the Helsinki Committee, researchers analyzed anonymous EMR data from Maccabi Healthcare Services, and trained machine learning models using common lab tests and basic demographic information.

With five different algorithms trained and tested, the study showed a promising framework for using machine learning to detect patients at risk for celiac disease.

The program is part of a national effort to alleviate the Israeli doctor shortage. (Illustrative). (credit: INGIMAGE)
The program is part of a national effort to alleviate the Israeli doctor shortage. (Illustrative). (credit: INGIMAGE)
The approach could be integrated into healthcare practices where comprehensive EMR systems are in place, even potentially evolving into a prescreening method to flag patients for further evaluation.

Better healing and reduced symptoms

“Early identification of celiac disease can significantly improve patient outcomes, as those diagnosed earlier often experience better intestinal healing and reduced symptoms, whereas delayed diagnosis is linked to persistent health issues despite adhering to a gluten-free diet,” said Dr. Amir Ben Tov, the pediatric gastroenterologist and senior clinical researcher at KSM Research and Innovation Center.