Estimating Blood Pressure Measurements through an Electro-vascular-gram: A pilot study

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Signal Processing

Background:

• Accurate blood pressure (BP) measurement is crucial for managing cardiovascular disease.
• Current BP monitoring methods have limitations, necessitating innovative approaches.
• Non-invasive or minimally invasive techniques for continuous BP assessment are highly desirable.

Purpose of the Study:

• To investigate the feasibility of using electro-vascular-gram (EVG) signals for predicting arterial blood pressure.
• To develop and validate a signal processing and machine learning model for BP estimation.
• To assess the accuracy of the proposed method in a preclinical setting.

Main Methods:

• Acute experiments were performed on male rats under urethane anesthesia.
• A cuff electrode was surgically implanted around the carotid artery to measure EVG signals.
• Direct BP measurements were obtained via catheterization of the contralateral carotid artery for ground truth.
• A convolutional neural network (CNN) was employed to process EVG signals and predict systolic (SYS) and diastolic (DIA) BP.

Main Results:

• The CNN model successfully predicted SYS and DIA BP values with an error less than 5mmHg.
• Prediction accuracy for SYS BP ranged from 45-98% across testing datasets.
• Prediction accuracy for DIA BP ranged from 52-97% across testing datasets.
• Preliminary data indicates the feasibility of continuous BP measurement using EVG signals.

Conclusions:

• The electro-vascular-gram (EVG) signal, measured by a cuff electrode, shows potential for non-invasive blood pressure monitoring.
• Convolutional neural networks can effectively process EVG signals for accurate BP prediction.
• This preliminary study suggests a viable method for continuous arterial BP assessment, aiding cardiovascular health management.