Low perfusion algorithm used in wearable oximeter and hardware acceleration

This paper proposes a simple and efficient algorithm named “fundamental frequency detection and narrow bandpass filtering” to calculate SpO2 in wearable oximeter when it is in low perfusion condition. The two-stage pipeline structure is used to accelerate computing when the algorithm is implemented by hardware logic circuit. Through the FPGA verification, the system can output real-time, continuous calculation, and extract pulse wave from as low as 0.1% perfusion signal then accurately calculate SpO2 value with an error less than 3%. It can entirely satisfy the medical demands.