Correlation algorithm and sampling techniques for estimating the signal-to-noise ratio of the electrocardiogram

Area of Science:

• Signal Processing
• Biomedical Engineering
• Noise Analysis

Background:

• Accurate signal-to-noise ratio (SNR) estimation is crucial for analyzing very low frequency (VLF) signals.
• VLF signals are often contaminated by various noise types, including white and flicker noise, complicating analysis.
• Existing methods may struggle with the specific challenges posed by VLF signals and combined noise sources.

Purpose of the Study:

• To propose and validate a novel algorithm for estimating the SNR of VLF signals.
• To address the challenge of VLF signal analysis in the presence of white and flicker noise.
• To introduce innovative sampling techniques for enhanced correlation analysis.

Main Methods:

• Development of an algorithm based on correlation techniques for SNR estimation.
• Implementation of novel sampling strategies to convert a single continuous signal into two time series.
• Validation of the algorithm using simulated datasets and real-world electrocardiogram (ECG) data from ten patients.

Main Results:

• The proposed algorithm successfully estimates the SNR of VLF signals contaminated by white and flicker noise.
• The sampling techniques enable the effective application of cross-correlation functions to the analyzed signals.
• The algorithm demonstrated reliable performance on both simulated data and clinical ECG recordings.

Conclusions:

• The developed correlation-based algorithm provides a robust method for SNR estimation in VLF signals.
• The proposed sampling technique is effective for analyzing noisy VLF data, including biomedical signals.
• This approach offers a valuable tool for researchers and practitioners working with low-frequency signal analysis.