Adaptive Spectro-Temporal Filtering for Electrocardiogram Signal Enhancement

Area of Science:

• Biomedical Engineering
• Signal Processing

Background:

• Portable electrocardiogram (ECG) devices are increasingly used for fitness and health monitoring.
• Movement and muscle artifacts significantly contaminate ECG signals, leading to inaccurate heart rate and heart rate variability (HRV) analysis.

Purpose of the Study:

• To develop and evaluate a novel denoising method for enhancing ECG signals contaminated by artifacts.
• To improve the reliability of ECG-based analyses for applications like fitness tracking and athletic performance.

Main Methods:

• Proposed an adaptive spectro-temporal filtering algorithm utilizing modulation spectral signal representation.
• Tested the algorithm on synthetic ECG signals with varying noise levels and real-world noisy recordings.
• Compared performance against a state-of-the-art wavelet-based denoising method.

Main Results:

• The proposed method achieved a 61.8% improvement in signal-to-noise ratio.
• Demonstrated a three-fold reduction in average heart rate measurement error.
• Showed a 15% reduction in HRV measurement error compared to the benchmark algorithm, especially in noisy conditions.

Conclusions:

• The adaptive spectro-temporal filtering method offers superior ECG denoising capabilities, particularly for noisy signals.
• This algorithm is a promising tool for enhancing the accuracy of ECG-based fitness and athletic monitoring applications.