Nocturnal Blood Pressure Estimation from Sleep Plethysmography Using Machine Learning

We demonstrated a method to obtain within 5mm/Hg MAE quality blood pressure measurements using a random forest ML algorithm trained on finger-tip photoplethysmography data. This work opens the door for seamless high frequency sampling of nocturnal blood pressure using wearable sensing that is comfortable. Note that current implementations like those used by Samsung using oscillometry require user action to obtain a reading.