Efficient compressive sensing of ECG segments based on machine learning for QRS-based arrhythmia detection.

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Arrhythmia telemonitoring is crucial for cardiac patient care.
• Efficient processing of electrocardiogram (ECG) data is essential for real-time analysis.
• Current methods can be computationally intensive, limiting long-term monitoring capabilities.

Purpose of the Study:

• To propose a novel and efficient method for telemonitoring arrhythmia.
• To reduce computational load in ECG analysis for remote monitoring.
• To enable faster detection of QRS complexes and their absence in ECG segments.

Main Methods:

• Development of a telemonitoring method based on QRS complex analysis.
• Efficient computation of Sum of Absolute Differences (SAD) and Maximum of Absolute Differences (MAD) features from ECG segments.
• Utilizing computed features for wireless transmission and QRS complex absence determination at the receiver.

Main Results:

• The proposed method reduces computation time by approximately 7.4% for long-term telemonitoring.
• Detection of QRS complex absence is achieved in approximately 7 milliseconds on standard hardware.
• Avoids computationally expensive signal reconstruction for segments without QRS complexes.

Conclusions:

• The novel method offers an efficient approach to telemonitoring arrhythmia.
• Significant reduction in computation time enhances the feasibility of long-term remote cardiac monitoring.
• The method provides a rapid and reliable way to detect the absence of QRS complexes, aiding in arrhythmia diagnosis.