A Low-Power Predictive Sampling PPG Sensor

Photoplethysmography (PPG) sensors provide accurate measurements of vital signs, including heart rate (HR), HR variability (HRV), and oxygen saturation (SpO
2
). PPG signal detection needs light-emitting diodes with sufficient brightness, consuming several mW in continuous monitoring. We present a PPG sensor that adaptively predicts peaks and valleys (PAVs), i.e., the maximum and minimum amplitudes of the signal, and extracts HR, HRV, SpO
2
, systolic, diastolic, and cardiac cycles with exceptionally low power consumption. The sensor activates only when PAVs occur, resulting in a power reduction of >87%, while maintaining a mean absolute error (MAE) of <0.5 beats per minute for HR, <1% for SpO
2
, and <3 ms for HRV, systolic, diastolic, and cardiac cycles. Our prototype was tested on the fingertip, wrist, arm, chest, neck, earlobe, and forehead of a 29-year-old subject.