Israeli-developed algorithm predicts gestational diabetes

Segal and his colleagues applied a machine learning method to Clalit's records on some 450,000 pregnancies in women who gave birth in the years 2010 to 2017.

A person receives a test for diabetes during Care Harbor LA free medical clinic in Los Angeles, California September 11, 2014 (photo credit: MARIO ANZUONI/REUTERS)
A person receives a test for diabetes during Care Harbor LA free medical clinic in Los Angeles, California September 11, 2014
(photo credit: MARIO ANZUONI/REUTERS)
A new computer algorithm developed by researchers at the Weizmann Institute of Science in Rehovot can predict which women are at a high risk of gestational diabetes in the early stages of pregnancy or even before it has occurred, the institute said in a press release Monday afternoon.
The study analyzed data on nearly 600,000 pregnancies available from Israel's largest health insurance provider, Clalit Health Services, the Weizmann Institute of Science said. According to the institute, the algorithm may help prevent gestational diabetes using nutritional and lifestyle changes.
Gestational diabetes occurs in three to nine percent of all pregnancies and is characterized by high blood sugar that develops during pregnancy in non-diabetic women. While often disappearing after the pregnancy, it may be dangerous for both the mother and the baby. Typically, gestational diabetes is diagnosed between the 24th and 28th weeks of pregnancy.
"Our ultimate goal has been to help the health system take measures so as to prevent diabetes from occurring in pregnancy," said Prof. Eran Segal of the Computer Science, Applied Mathematics and Molecular Cell Biology departments.
Segal and his colleagues applied a machine learning method to Clalit's records on some 450,000 pregnancies in women who gave birth in the years 2010 to 2017. Gestational diabetes had been diagnosed by glucose tolerance testing in about four percent of them.
After analyzing a 2,000-parameter dataset, the algorithm managed to identify nine parameters – including age, BMI, family history and results of glucose tolerance testing from previous pregnancies – that were sufficient to identify those at risk of developing gestational diabetes.
The researchers reassured the efficiency of the algorithm's predictions that were based on these parameters by applying them to records on around 140,000 additional pregnancies that were not included in the initial analysis, validating the study's findings.
The research was led by graduate students Nitzan Shalom Artzi, Dr. Smadar Shilo and Hagai Rossman from Eran Segal’s lab at the Weizmann Institute, who collaborated with Prof. Eran Hadar, Dr. Shiri Barbash-Hazan, Prof. Avi Ben-Haroush and Prof. Arnon Wiznitzer of Rabin Medical Center in Petah Tikva; and Prof. Ran D. Balicer and Dr. Becca Feldman of Clalit.