Novel Blood Pressure Waveform Reconstruction from Photoplethysmography using Cycle Generative Adversarial Networks
This work addresses the need for accurate, non-invasive blood pressure monitoring in free-living conditions for individuals with chronic diseases like hypertension, representing a domain-specific incremental advance.
The paper tackled the problem of non-invasive continuous blood pressure monitoring by proposing a CycleGAN-based method to reconstruct ambulatory blood pressure waveforms from photoplethysmography signals, achieving up to 2x improvement in estimation accuracy over state-of-the-art approaches.
Continuous monitoring of blood pressure (BP)can help individuals manage their chronic diseases such as hypertension, requiring non-invasive measurement methods in free-living conditions. Recent approaches fuse Photoplethysmograph (PPG) and electrocardiographic (ECG) signals using different machine and deep learning approaches to non-invasively estimate BP; however, they fail to reconstruct the complete signal, leading to less accurate models. In this paper, we propose a cycle generative adversarial network (CycleGAN) based approach to extract a BP signal known as ambulatory blood pressure (ABP) from a clean PPG signal. Our approach uses a cycle generative adversarial network that extends theGAN architecture for domain translation, and outperforms state-of-the-art approaches by up to 2x in BP estimation.