ECG feature extraction using principal component analysis for studying the effect of diabetes

Area of Science:

• Biomedical Engineering
• Healthcare Technology
• Cardiology

Background:

• Electrocardiogram (ECG) analysis is crucial for assessing cardiac health.
• Diabetes is a prevalent chronic illness with significant implications for cardiovascular well-being.
• Accurate ECG parameter identification is vital for research and clinical practice.

Purpose of the Study:

• To develop and present an algorithm for 12-lead ECG feature extraction using Principal Component Analysis (PCA).
• To estimate diabetes mellitus (DM)-related ECG parameters, including corrected QT interval (QTc), QT dispersion (QTd), P wave dispersion (PD), and ST depression (STd).
• To investigate the relationship between diabetes and cardiac health through advanced ECG analysis.

Main Methods:

• Data acquisition using the XBio Aqulyser unit.
• Preprocessing steps including Fast Fourier Transform (FFT) for baseline wander removal and wavelet transform for signal denoising.
• Principal Component Analysis (PCA) for R-wave extraction, followed by window-based methods for other wave identification.
• Estimation of diabetes-specific ECG parameters (QTc, QTd, PD, STd) from extracted features.

Main Results:

• Successful extraction of R-waves and other ECG components using the proposed PCA-based algorithm.
• Estimation of key diabetes-related cardiac parameters from the processed ECG signals.
• Demonstrated the feasibility of the algorithm in analyzing ECG data from 25 diabetic patients.

Conclusions:

• The developed PCA-based algorithm provides an effective method for ECG feature extraction and the estimation of diabetes-related parameters.
• This approach holds potential for improving the assessment of cardiac health in diabetic individuals.
• Further research with larger datasets is warranted to validate the clinical utility of this ECG analysis technique.