Key result
A CycleGAN-based deep learning model reconstructed ambulatory blood pressure waveforms from photoplethysmography signals with a mean absolute error of 2.89 mmHg for systolic and 3.22 mmHg for diastolic blood pressure.
Why the study?
Existing non-invasive methods combining PPG and ECG signals fail to reconstruct the complete blood pressure waveform, resulting in less accurate models for continuous monitoring.
Does a CycleGAN-based approach improve the accuracy of blood pressure waveform reconstruction from PPG signals compared to existing methods?
Does a CycleGAN-based approach improve the accuracy of blood pressure waveform reconstruction from PPG signals compared to existing methods?
A novel CycleGAN-based deep learning model can accurately reconstruct complete ambulatory blood pressure waveforms from non-invasive PPG signals, significantly improving estimation accuracy over existing methods.
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.
No takes yet. Share an insight, caveat, or question.
Mehrabadi et al. (2022) studied Blood pressure monitoring (n=92). Cycle Generative Adversarial Network (CycleGAN) vs. State-of-the-art blood pressure estimation approaches was evaluated on Mean absolute error (MAE) of systolic and diastolic blood pressure estimation in cross-subject evaluation (p=<.001). A CycleGAN-based deep learning model reconstructed ambulatory blood pressure waveforms from photoplethysmography signals with a mean absolute error of 2.89 mmHg for systolic and 3.22 mmHg for diastolic blood pressure.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: