Important ECG diagnosis-aiding indices of ventricular septal defect children with or without congestive heart failure

Area of Science:

• Cardiology
• Biomedical Engineering
• Medical Informatics

Background:

• Electrocardiogram (ECG) analysis is crucial for diagnosing pediatric heart conditions.
• Traditional analysis focuses on RR intervals, but new metrics may offer improved diagnostic power.
• Ventricular Septal Defect (VSD) and Congestive Heart Failure (CHF) require accurate early detection.

Purpose of the Study:

• To statistically evaluate the diagnostic utility of novel PR' and RT intervals alongside conventional RR intervals in pediatric ECGs.
• To develop and compare classification methods for distinguishing between normal children, and those with VSD with or without CHF.
• To propose advanced logistic regression models for improved diagnostic accuracy.

Main Methods:

• Statistical analysis of PR', RR, and RT intervals from ECG data.
• Application of quadratic classification rules and search for optimal classification vectors.
• Development of an automated ECG interval detection algorithm with outlier detection.
• Implementation and evaluation of logistic regression models and receiver operating characteristic (ROC) analysis.

Main Results:

• Specific statistics derived from PR', RR, and RT intervals serve as significant diagnosis-aiding indices.
• The proposed logistic regression models demonstrate superior performance compared to linear and quadratic logistic models.
• The automated algorithm accurately measures ECG intervals, enhancing data analysis reliability.

Conclusions:

• PR', RR, and RT intervals provide valuable information for diagnosing pediatric VSD and CHF.
• Advanced statistical and machine learning models based on these intervals improve diagnostic accuracy.
• Automated ECG analysis tools are essential for reliable and efficient clinical application.