A new approach for blood pressure estimation based on phonocardiogram

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Medical Devices

Background:

• Continuous, non-invasive blood pressure (BP) monitoring is crucial for critical care patients.
• Existing cuffless methods like pulse transit time (PTT) are sensitive and cumbersome for ambulatory use.
• Phonocardiogram (PCG) signals offer a potential, less invasive data source for BP estimation.

Purpose of the Study:

• To develop and validate a novel method for estimating blood pressure (BP) using phonocardiogram (PCG) signals.
• To explore the correlation between pulse transit time (PTT) and diastolic duration (S21) for BP estimation.
• To assess the feasibility of an artificial neural network (ANN) approach for cuffless BP monitoring.

Main Methods:

• An artificial neural network (ANN) was developed, utilizing systolic duration, diastolic duration, heart rate, sex, height, and weight as input features.
• The ANN was trained to estimate mean blood pressure, subsequently deriving systolic and diastolic pressures.
• The proposed PCG-based method was evaluated on 37 subjects against a commercial blood pressure device.

Main Results:

• The ANN model demonstrated satisfactory accuracy in estimating systolic and diastolic pressures.
• The mean error for systolic pressure estimation was $$ mmHg, and for diastolic pressure was $$ mmHg.
• These error margins are comparable to the Association for the Advancement of Medical Instrumentation (AAMI) standard of $$ mmHg.

Conclusions:

• The study validates the feasibility of estimating blood pressure using PCG signals.
• The proposed ANN-based method offers a promising, non-invasive, and cuffless approach for continuous BP monitoring.
• This technique has the potential to improve patient care, particularly in critical and ambulatory settings.