Diagnosis of OSA by Machine Learning and Signal Processing Using PPG Signals and HRV Parameters

The current study reviews recent advancements in the processing and acquisition of PPG signals, such as the integration of accelerometers to produce hypnograms for both good and unhealthy sleep,as well as the method for identifying breathing disorders during sleep (SDB). The thorough application of machine learning and signal processing approaches to diagnose OSA using PPG signal and HRV parameters has been described. Additionally, there are numerous suggested models and potential PPG medical applications. The downsides of PPG have also been acknowledged, and technical solutions to these issues have been considered. These limitations include oximetry signal failure, motion artifacts, and signal processing.