Optimized Performance Pulse Oximeter Based on the MAX30102 Commercial Sensor

Abstract

Pulse oximeters are devices that use the photoplethysmography technique to estimate oxygen saturation in blood and heart rate. The MAX30102 is a sensor for reflective photoplethysmography with signal-conditioning and digitalization stages embedded in a single chip that facilitate its implementation in wearable devices. However, there are limitations and external factors that affect its performance in a significant way. This paper suggests a pulse oximeter based on the MAX30102 whose performance has been optimized through the design of algorithms adapted to the signals of this particular sensor. The performance of two heart rate measurement algorithms is compared, one based on a pulse counter and the other on the Fast Fourier Transform (FFT). The proposal covers from the algorithms design to the manufacture of a functional prototype tested with volunteers. Results indicate an accuracy of ±1.39% for the measurement of oxygen saturation and ±2.04 bpm for heart rate. The Bland-Altman analysis of the pilot test results indicate that prototype measurements are comparable to those of a high-end oximeter taken as reference.