Cuffless Hypertension Detection using Swarm Support Vector Machine Utilizing Photoplethysmogram and...

Area of Science:

• Biomedical Engineering
• Cardiovascular Health
• Artificial Intelligence in Medicine

Background:

• Hypertension poses significant health risks, necessitating early detection to prevent severe complications.
• Timely identification of hypertension events is crucial for effective management and improved patient outcomes.

Purpose of the Study:

• To investigate a novel, cuff-free strategy for hypertension detection.
• To utilize bioelectric signals, specifically Electrocardiogram (ECG) and Photoplethysmogram (PPG), for non-invasive blood pressure monitoring.
• To develop and optimize a Swarm-based Support Vector Machine (SSVM) algorithm for accurate hypertension detection.

Main Methods:

• Collected ECG and PPG data from normal and hypertensive participants from the MIMIC database.
• Extracted key parameters including Pulse Arrival Time (PAT) and PPG signal derivatives.
• Employed an SSVM algorithm, optimized with Quantum Delta-potential-well Particle Swarm Optimization (QDPSO), to analyze extracted parameters.

Main Results:

• The proposed SSVM strategy achieved a 96% performance rate across F1-score, accuracy, sensitivity, and specificity.
• The developed system demonstrated superior performance compared to other tested methods.
• The study successfully developed a functional cuff-free hypertension monitoring system.

Conclusions:

• The SSVM algorithm, utilizing ECG and PPG parameters, provides an acceptable method for hypertension detection.
• Integrating both ECG and PPG signals yields better hypertension detection performance than using PPG alone.
• The research validates the potential of non-invasive bioelectric signal analysis for hypertension management.