Science & Enterprise logo
Science for business people. Enterprise for scientists.

AI-Aided System to Identify Diabetes Complication Risks

Diabetes blood glucose test
(Amanda Mills, CDC.gov)

4 Nov. 2022. Researchers at University of Houston are creating a system aided by algorithms for primary care physicians to better identify people at risk for complications from diabetes. A team led by Winston Liaw, a data scientist and public health researcher in the university’s medical school, is developing the Diabetes Complication Severity Index progression tool, funded by a grant from American Board of Family Medicine.

Diabetes is a chronic condition where the pancreas creates no or not enough insulin to process the sugar glucose in the blood stream, providing energy for cells in the body. In type 2 diabetes, which accounts for at least 90 percent of all diabetes cases, the pancreas produces some but not enough insulin, or the body cannot process insulin. Centers for Disease Control and Prevention says some 37.3 million people in the U.S. have diabetes, with 20 percent of that population unaware of their condition. Moreover, another 96 million or one of three American adults have prediabetes, a condition where the body’s sugar processing system begins to break down and glucose levels are higher than normal, but not yet at type 2 diabetes.

For many people, diabetes leads to complications including chronic kidney disease, heart disease, vision impairment, nerve damage, and slow-healing skin wounds on lower limbs that sometimes result in amputations. One way that physicians and insurance companies today try to identify patients at higher risk of complications from diabetes is the Diabetes Complication Severity Index, or DCSI, a standard scale that quantifies long-term effects of diabetes on seven systems in the body.

Wider range of social and environmental factors

The Houston researchers note, however, that DCSI calculates diabetes complication risks only at a single point in time. The team proposes enhancing DCSI to provide earlier warnings of people at higher risk of diabetes complications, giving patients and primary care physicians more options in dealing with those risks. Liaw and colleagues are writing machine learning algorithms, a form of artificial intelligence, bringing in a wider range of social and environmental factors affecting a person’s health, such as education level, employment status, and food security.

The researchers say the algorithms will be written with the university’s health system sciences institute, funded by insurance company Humana, and based in part on data sets with claims and health records provided by Humana. The team plans to test the system with Prime Registry, an independent database of primary care health records.

The DCSI progression index project is funded by an award earlier this year from American Board of Family Medicine, as part of that group’s efforts to promote machine learning and artificial intelligence in family practice. A university spokesperson tells Science & Enterprise in an email the grant provides some $500,000 over four years.

The researchers expect the DCSI progression index will provide primary care physicians with insights into risks of diabetes complications by patients, enabling earlier actions to reduce those complications. “Our long-term goal is to help clinicians become more proactive and less reactive when treating diabetes,” says Liaw in a University of Houston statement. “By leveraging the capabilities of artificial intelligence and machine learning, we can more effectively connect at-risk individuals with interventions before they become sicker.”

More from Science & Enterprise:

We designed Science & Enterprise for busy readers including investors, researchers, entrepreneurs, and students. Except for a narrow cookies and privacy strip for first-time visitors, we have no pop-ups blocking the entire page, nor distracting animated GIF graphics. If you want to subscribe for daily email alerts, you can do that here, or find the link in the upper left-hand corner of the desktop page. The site is free, with no paywall. But, of course, donations are gratefully accepted.

[wpedon id=”42724″ align=”center”]

*     *     *