Cuff-less blood pressure measurement based on hybrid feature selection algorithm and multi-penalty regularized...

Area of Science:

• Biomedical Engineering
• Cardiovascular Health
• Medical Diagnostics

Background:

• Hypertension is a major risk factor for cardiovascular disease, often undetected until significant organ damage occurs.
• Frequent blood pressure (BP) monitoring is essential for early detection and management, but cuff-based methods are inconvenient for daily use.
• Cuffless BP measurement offers a convenient alternative for continuous monitoring.

Purpose of the Study:

• To develop and validate a novel cuffless blood pressure measurement technique.
• To utilize electrocardiogram (ECG) and photoplethysmography (PPG) signals for BP estimation.
• To improve the accuracy and convenience of BP monitoring for hypertension management.

Main Methods:

• Feature extraction from ECG and PPG signals.
• Hybrid feature selection algorithms combined with a generalized penalty-based regression technique.
• Training and testing the model on the MIMIC-II database.
• Comparison with ordinary linear regression and other contemporary cuffless techniques.

Main Results:

• Significant reduction in Root Mean Square Error (RMSE): 11.2 to 5.6 mmHg for systolic BP (SBP) and 12.7 to 6.69 mmHg for diastolic BP (DBP).
• Mean Absolute Error (MAE) achieved: 4.91 mmHg for SBP and 5.77 mmHg for DBP.
• The proposed method meets the British Hypertension Society (BHS) standard of grade B for BP measurement accuracy.

Conclusions:

• The proposed cuffless BP measurement technique demonstrates high accuracy and reliability.
• This method offers a promising, convenient alternative to traditional cuff-based BP monitoring.
• The findings support the potential of using ECG and PPG for effective hypertension management.