Health Medical

Medial EarlySign’s New Model Can Detect Prediabetic Patients at Greater Risk of Diabetes

Recent research was conducted to check the efficacy of the Pre2D-Flag AlgoMarker algorithmic model provided by Medial EarlySign. This new model is capable of diagnosing a prediabetic patient who has a greater risk of diabetes progression. The company has publicized the findings of the study on January 22, 2020, presented in Diabetes/Metabolism Research and Reviews.

Medial EarlySign is a leading company known for its AI solutions. The machine learning-based solutions offered by the Medial EarlySign facilitate healthcare providers in the prevention and early diagnosis of various high-burden diseases like diabetes. 

This retrospective study was conducted via cooperation between the Medical division of MHS (Maccabi Healthcare Services) and the Diabetes Unit at Hadassah Hebrew University of Jerusalem. 

The logistic-regression models – the alternatives for this new technology weren’t found to be as effective as Pre2D-Flag AlgoMarker provided by EarlySign. The latter was observed to be better at predicting diabetes. Unlike logistic-regression models, this new AI technology has the ability to detect even a slight relationship between multiple variables. 

Read More: New Non-Fibrillating Insulin Compound Showed Promising Results In Diabetes Treatment

Diabetes is a chronic illness that results in uncontrolled levels of glucose in the blood. Avivit Cahn is the study’s leading author and is working in Hadassah University Hospital as a Senior Endocrinologist. Nearly 727 billion dollars – 12 percent of the global health expenditure, is used for dealing with diabetes and the complications associated with it. 

According to Avivit Cahn, the continuous increase in diabetes prevalence has become a global health concern. Traditional methods to determine the risk of diabetes in an individual including observation of clinical signs and symptoms and biochemical testing are only helpful to a limited extent.  

By using electronic health records, this novel AI (machine learning) model can identify prediabetic patients who have a greater diabetes risk. The early diagnosis of diabetes and timely interventions targeted at the high-risk group instead of all individuals with prediabetes can lead to a considerable reduction in the overall burden of this global health problem. 

Read More: This Diabetes Drug May Treat recurring Miscarriages in Women

Pre2D-Flag AlgoMarker by EarlySign had been instructed on the basis of the primary care database of the U.K. National Health Service and the data obtained from The Health Improvement Network (THIN) database. This model was further validated based on the datasets from the Israeli Maccabi Healthcare Services and the Canadian AppleTree.

With the help of the THIN database, the data of 852,454 people diagnosed with prediabetes was extracted for the purpose of the study. The novel AI model utilizing 69 variables was applied to these individuals, allowing the examination of the last 10 years’ data and using this data to determine the diabetes risk in the coming year.

The findings indicated that this novel system is a better way for diabetes prediction even if the historical data consists of 3 to 5 years. This AI technology can also help in detecting diabetes risk for the next five years.  

This breakthrough technology can make a change in diabetes prevalence by facilitating clinicians in early detection and more targeted and timely interventions for diabetes. The flexibility and continuity in this recent AI model also allow health professionals to use them in different settings. 

About the author

Khadija Ahmad

Khadija is a highly talented author at Spark Health MD, who uses her high-end creativity to deliver best-in-class content regarding her research work to facilitate the reader as well as raise their knowledge about Health And Physical Education through her interesting articles containing informative as well as interactive content.

Add Comment

Click here to post a comment

Your email address will not be published. Required fields are marked *

error: Content is protected !!