Continuous Cuffless Non-Invasive Blood Pressure Detection from ECG and PPG Signals Using Artificial Neural Network

Hypertension, a primary cause of cardiovascular diseases (CVD), necessitates real-time and continuous blood pressure (BP) monitoring for accurate diagnosis and treatment. However, existing methods for continuous BP measurement are often invasive and painful, while non-invasive methods using cuffs lack convenience for continuous monitoring. This study proposes a continuous, non-invasive cuffless BP detection method using electrocardiogram (ECG) and photoplethysmogram (PPG) signals. By applying an Artificial Neural Network (ANN), we achieved accurate estimations of systolic (SBP) and diastolic (DBP) blood pressure. The model, developed with the MIMIC II database, demonstrated a Mean Absolute Error (MAE) of ±13.89 mmHg for SBP and ±5.5 mmHg for DBP. Comparisons with conventional methods showed an average error of 11.2 mmHg for SBP and 9.75 mmHg for DBP, validating its effectiveness for noninvasive, continuous BP monitoring without causing pain or risk of injury.