Improved late potential analysis in frequency domain

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Arrhythmogenic right ventricular disease (ARVD) is linked to late potentials (LP).
• Traditional time-domain analysis has limitations.
• Frequency-domain analysis of ECG signals is promising for ARVD detection.

Purpose of the Study:

• To enhance the identification of late potentials (LP) in patients with arrhythmogenic right ventricular disease (ARVD).
• To improve the reproducibility of frequency-domain ECG analysis for ARVD.
• To develop a reliable method for classifying ARVD patients and healthy subjects.

Main Methods:

• Spectral mapping of the ECG using Fast Fourier Transform (FFT) on ST segments.
• Development of an algorithm to improve QRS endpoint identification reproducibility.
• Application of an optimal QRS filter and Hilbert transform for enhanced normality factor estimation.

Main Results:

• The novel algorithm increased LP identification in ARVD patients from 47% to 88% for those with sustained ventricular tachycardia.
• LP identification rose from 18% to 64% in ARVD patients without sustained ventricular tachycardia.
• High sensitivity for ARVD and sustained ventricular tachycardia was achieved using cluster analysis.

Conclusions:

• The proposed technique significantly improves late potential detection in arrhythmogenic right ventricular disease.
• Enhanced QRS endpoint identification and signal processing increase diagnostic reliability.
• This method offers a more accurate tool for identifying ARVD and predicting ventricular tachycardia risk.