Classification of cardiac arrhythmias using fuzzy ARTMAP

Area of Science:

• Cardiology
• Biomedical Engineering
• Artificial Intelligence in Medicine

Background:

• Cardiac arrhythmias are a significant cause of morbidity and mortality.
• Accurate classification of arrhythmias like premature ventricular contractions (PVCs) is crucial for effective treatment.
• Electrocardiogram (ECG) signals contain vital information for arrhythmia detection.

Purpose of the Study:

• To investigate the efficacy of fuzzy adaptive resonance theory mapping (ARTMAP) for classifying cardiac arrhythmias.
• To differentiate between normal and abnormal premature ventricular contraction (PVC) conditions using ECG data.
• To develop a robust automated system for cardiac arrhythmia classification.

Main Methods:

• Extraction of QRS complexes from ECG data.
• Application of bandpass filtering, scaling, and Hamming windowing to QRS segments.
• Generation of two linear predictive coding (LPC) coefficients using Burg's maximum entropy method.
• Utilization of LPC coefficients and mean-square values as features for a fuzzy ARTMAP neural network.

Main Results:

• The fuzzy ARTMAP neural network achieved high classification performance.
• Specificity for classifying cardiac arrhythmias exceeded 99%.
• Sensitivity for classifying cardiac arrhythmias reached 97%.

Conclusions:

• Fuzzy ARTMAP is a highly effective tool for classifying cardiac arrhythmias, particularly PVCs.
• The developed method demonstrates excellent accuracy in distinguishing normal and abnormal cardiac rhythms.
• This approach holds promise for automated, real-time cardiac arrhythmia monitoring and diagnosis.