Early Detection of Cardiovascular Disease Using Photoplethysmography (PPG) Signal Analysis

Abstract

Photoplethysmography (PPG) signals have gained prominence in clinical diagnostics for their non-invasive, cost-effective, and user-friendly applications in detecting cardiovascular diseases (CVDs). This study leverages machine learning techniques to enhance the accuracy of CVD detection from PPG data, addressing critical risk factors such as hypertension and stress, which significantly contribute to elevated blood pressure and, consequently, to cardiovascular disorders. The use of PPG provides a reliable approach for identifying cardiovascular anomalies by monitoring essential parameters like blood pressure and heart rate. In this work, we employ both machine learning and deep learning, specifically neural networks, to assist clinicians in diagnosing CVD, achieving a high accuracy rate of 98% on the PPG-BP dataset. The findings demonstrate the potential of PPG signals combined with advanced algorithms to support early diagnosis and personalized treatment, ultimately reducing mortality
rates associated with cardiovascular diseases.