Author ORCID Identifier

https://orcid.org/0009-0004-0630-8978

Semester

Spring

Date of Graduation

2026

Document Type

Thesis

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Lane Department of Computer Science and Electrical Engineering

Committee Chair

Nima Karimian

Committee Member

Kevin Bandura

Committee Member

Muhammad Choudhry

Abstract

Abstract
Hyperglycemia Detection from Single-Lead ECG using a Hybrid CNN & Transformer Model
Adam Ogunjembola

Diabetes Mellitus is known as high blood glucose. This high blood glucose level happens when the body has a problem with producing or using insulin. Insulin is a very important hormone that the pancreas makes to control how much glucose gets into the bloodstream and cells. Diabetes Mellitus has an effect on the body if it is not treated, such as damaging the blood vessels and nerves which can lead to stroke, kidney failure, heart attack and permanent loss of vision. Since people with diabetes need to measure their glucose levels in order to determine how much insulin or medicine they need to take, reliable and convenient glucose measurement is crucial in diabetes treatment. Diabetes affects over 40 million Americans. For many of these people, frequent blood glucose testing becomes routine as part of their daily care. Traditional glucose monitoring methods include taking a drop of blood from a finger prick to be measured by a blood glucose meter. This process can be intrusive, painful and inconvenient when multiple tests are required or continuous monitoring is needed. Thus, this thesis propose a non-invasive, cheaper, less painful, and continuous method of glucose monitoring. Electrocardiograms have been previously considered to be able to measure about glucose levels in the body because when blood glucose levels rise, it affects the heart rate. In this project, we took ECG signals and created a hybrid model to process them. We used a CNN, transformer, and squeeze-excitation to determine the patterns in the ECG that corresponded to high blood glucose. We cleaned the ECG signal to remove noise and extracted and labelled features from the data. This then led us to create a hybrid model of CNN and transformer which yielded an AUC of 95%, sensitivity of 84%, and specificity of 90%. These results indicate that there is information in ECG signals about blood glucose levels and can be used with wearable devices.

Download

Share

COinS