Photoplethysmography-Based Method for Automatic Detection of Premature Ventricular Contractions

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Signal Processing

Background:

• Premature ventricular contractions (PVCs) are common arrhythmias.
• Current detection methods often rely on electrocardiography (ECG), which can be obtrusive.
• Photoplethysmogram (PPG) signals offer a potential non-invasive alternative for monitoring cardiac activity.

Purpose of the Study:

• To develop and validate a novel method for detecting PVCs using PPG signals.
• To assess the performance of artificial neural networks in classifying PVCs based on PPG features.
• To compare the efficacy of PPG-based PVC detection against traditional ECG-based methods.

Main Methods:

• Extraction of 6 key features from PPG signals, including pulse power and peak-to-peak intervals.
• Application of a sliding window approach for feature extraction and normalization against estimated heart rate.
• Utilizing artificial neural networks with linear and non-linear outputs for classification of PVCs.
• Training and testing on PhysioNet databases (MIMIC II and MIMIC).

Main Results:

• High sensitivity and specificity achieved for PVC detection: 92.4%/99.9% and 93.2%/99.9% for PVCs with and without observable PPG pulses, respectively.
• Successful differentiation between PVC types based on PPG signal characteristics.
• Demonstrated the feasibility of using PPG for reliable PVC detection.

Conclusions:

• The developed PPG-based method provides accurate and reliable detection of premature ventricular contractions.
• This approach offers a less invasive and potentially more accessible alternative to ECG-based monitoring for PVCs.
• The findings support the clinical utility of PPG signals in cardiovascular health monitoring.