PROJECT TITLE :

Predicting Quality of Overnight Glycaemic Control in Type 1 Diabetes Using Binary Classifiers

ABSTRACT:

Maintaining nocturnal blood glucose levels within the therapeutic range in type 1 diabetes might be difficult. Although semi-automatic devices to control insulin pump supply, such as low-glucose insulin suspension and the artificial pancreas, are becoming more common, their high cost and poor performance are preventing widespread use. As a result, a decision support system that assists persons with type 1 diabetes who are on numerous daily injections or insulin pump therapy in avoiding unwanted nighttime blood glucose variations (hyper- or hypoglycemic) is an appealing option. By evaluating commonly acquired data during the day-time period, we offer a unique data-driven approach to predict the quality of nightly glycaemic control in persons with type 1 diabetes in this work (continuous glucose monitoring data, meal intake and insulin boluses). The suggested method can forecast whether nocturnal blood glucose concentrations will stay within or beyond the desired range, allowing the user to take the necessary preventive measures (snack or change in basal insulin). On a publicly available clinical dataset, a variety of popular established Machine Learning methods for binary classification were assessed and compared for this purpose (i.e., OhioT1DM). Although there is no clearly superior classification system, this study shows that it is possible to predict the quality of overnight glycaemic control with moderate accuracy (AUC-ROC = 0.7) using regularly collected data in type 1 diabetes care.