Motion artifact cancellation for wearable photoplethysmographic sensor

Abstract

Photoplethysmography (PPG) is a non-invasive and unobtrusive technique to measure heart rate from the surface of the skin, by exposing a section of the skin to an LED light and measuring the changes in reflected light due to the blood pulsing through under the skin. However, PPG signals are highly susceptible to motion artifacts, and in order for this technique to be useful for heart rate measurements around-the-clock, a motion artifact cancellation mechanism must be implemented to recover the blood volume pulse (BVP) from the corrupted signal. Various digital signal processing (DSP) approaches for motion artifact cancellation have been attempted in the past, but a reliable around-the-clock PPG sensor is yet to be out on the market. This thesis outlines a novel, analog implementation of the motion artifact cancellation system, to evaluate the impact of sensor front-end improvements on the performance of motion artifact cancellation.