Electrocardiographic T-wave peak-to-end interval for hypoglycaemia detection

Area of Science:

• Cardiology
• Biomedical Engineering
• Machine Learning in Healthcare

Background:

• The T wave peak-to-end interval (TpTe) is a T wave morphology parameter.
• TpTe contains indicators for hypoglycemia detection.
• Existing detection methods often rely on corrected QT interval (QTc) and heart rate.

Purpose of the Study:

• To evaluate the corrected T wave peak-to-end interval (TpTe(c)) as an input for hypoglycemia detection.
• To compare the performance of machine learning models using TpTe(c) versus traditional parameters.

Main Methods:

• Utilized Support Vector Machine (SVM) and Fuzzy Support Vector Machine (FSVM) with Radial Basis Function (RBF).
• Compared classification systems with inputs of TpTe(c), QTc, and heart rate against systems using only QTc and heart rate.
• Assessed performance based on sensitivity, specificity, and accuracy.

Main Results:

• The inclusion of TpTe(c) alongside QTc and heart rate significantly improved hypoglycemia detection.
• The combined parameter set demonstrated superior sensitivity, specificity, and accuracy compared to traditional inputs alone.
• SVM and FSVM models showed effectiveness in classifying hypoglycemia based on ECG parameters.

Conclusions:

• Corrected T wave peak-to-end interval (TpTe(c)) is a valuable indicator for hypoglycemia detection.
• Integrating TpTe(c) with QTc and heart rate enhances the accuracy of hypoglycemia detection systems.
• Machine learning approaches, particularly SVM and FSVM, are effective tools for analyzing ECG data for metabolic disorder detection.