An algorithm for sleep apnea detection from single-lead ECG using Hermite basis functions.

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Sleep apnea is a common sleep disorder with significant health implications.
• Accurate detection of sleep apnea is crucial for timely diagnosis and treatment.
• Electrocardiogram (ECG) signals offer a non-invasive and accessible method for physiological monitoring.

Purpose of the Study:

• To develop and evaluate a novel methodology for sleep apnea detection using single-lead ECG signals.
• To assess the effectiveness of Hermite basis functions and R-R interval features for sleep apnea classification.
• To compare the performance of various machine learning classifiers for sleep apnea detection.

Main Methods:

• QRS complexes in ECG signals were approximated using Hermite basis functions.
• Features extracted included Hermite coefficients, R-R interval statistics, and QRS approximation error energy.
• Classification was performed using K-nearest neighbor (KNN), multilayer perceptron neural network (MLPNN), support vector machine (SVM), and least-square support vector machine (LS-SVM).

Main Results:

• Minute-by-minute classification achieved approximately 84% accuracy using LS-SVM with a Gaussian RBF kernel.
• Per-recording classification achieved approximately 97.14% accuracy using SVM and LS-SVM classifiers.
• The proposed method demonstrated comparable accuracy to existing techniques with reduced computational cost.

Conclusions:

• The proposed methodology effectively detects sleep apnea from single-lead ECG signals.
• The feature set derived from Hermite expansion and R-R intervals is robust for sleep apnea classification.
• This approach offers a computationally efficient and accurate alternative for sleep apnea detection.