Estimation of Continuous Arterial Blood Pressure Waveform from Photoplethysmogram in Terms of Loss Function Focus on Blood Pressure

Blood pressure (BP) monitoring is one of the most important factors in preventing cardiovascular disease. Existing blood pressure measurement methods have limitations in that they are inappropriate for real-time blood pressure monitoring. Recently, many studies have been conducted to estimate blood pressure using Photoplethysmogram (PPG), which has several advantages. Arterial blood pressure (ABP) reconstruction studies from PPG have attempted to overcome the limitations of existing studies. The ABP waveform can obtain the most accurate blood pressure and also contribute to the diagnosis and prevention of cardiovascular diseases. In this study, we propose a novel loss function that can be applied to a machine-learning or deep-learning model that reconstructs PPG into ABP. We confirmed the improvement in blood pressure estimation performance by applying the proposed loss function.