Matched Filtering for Heart Rate Estimation on Compressive Sensing ECG Measurements

Area of Science:

• Biomedical Engineering
• Signal Processing

Background:

• Long-term electrocardiogram (ECG) monitoring is crucial for diagnosing and tracking diseases.
• Compressive Sensing (CS) offers a low-complexity framework for ECG signal compression in wireless body sensor networks.
• Beat-to-beat intervals, derived from R-peak detection, are vital clinical information.

Purpose of the Study:

• To develop a heart rate estimation method for CS ECG recordings that bypasses signal reconstruction.
• To enable efficient, low-power, real-time monitoring of cardiac activity.

Main Methods:

• ECG signals are represented as CS linear measurements.
• QRS complex locations are estimated directly in the compressed domain using template correlation.
• The proposed method avoids computationally intensive signal reconstruction algorithms.

Main Results:

• The developed method achieves competitive performance compared to techniques applied to reconstructed ECG signals.
• Experimental validation on real ECG data confirms the efficacy of the compressed domain approach.

Conclusions:

• The proposed method offers a significant advantage for real-time, low-power applications by eliminating the need for ECG signal reconstruction.
• This technique facilitates efficient heart rate estimation from compressed ECG data, enhancing usability in wireless monitoring systems.