Advantages of Modeling Photoplethysmography (PPG) Signals using Variational Autoencoders

Photoplethysmography (PPG) signal analysis is an emerging field of research and has been applied to a variety of tasks like disease detection and blood pressure monitoring, and the signal waveforms, when properly deciphered, continue to expose increasing levels of detail about the physiology of a person. Variational Autoencoders (VAE) is a fundamental deep learning technique that is within the category of generative models in artificial intelligence. The transformative nature of VAEs enable a powerful approach of processing and interpreting PPG signals. In this paper, we propose a VAE model for PPG heart beats, called PPG-VAE (~1.67% MAE), and discuss its advantages and sample applications. We show how this model can help identify localized slope of a PPG Heart Beat (HB) Wave or pulse, remove localized high frequency noise, and generate new signal segments matching existing signal segment morphology. Finally, we conclude by describing how the model can be used for the compression of PPG signal data, as well as how the model’s utility for signal synthesis could be extended to provide granular control over the features of the created signal.