Hardware and Algorithmic Approaches to Combat Motion Artifacts in Photoplethysmographic Data

This chapter provides some historical context and the most recent hardware and algorithmic advances in detection, removal, and reconstruction of photoplethysmographic (PPG) data segments contaminated by motion artifacts (MA). In hardware, some recent advances in the development of multi-channel PPG sensors are highlighted, as they offer the possibility of obtaining a channel of PPG data with less MA, in combination with robust signal processing algorithms that can automatically choose the least MA-affected channel. Traditional algorithmic approaches to detect, remove, and reconstruct MA-contaminated PPG data segments are described. In addition, some of the promising new techniques for MA detection and reconstruction of the contaminated PPG data segments including machine and deep learning approaches are described. Finally, applications of PPG sensors beyond obtaining heart rate, respiratory rate, and oxygen saturation are explored. The new applications should motivate further advances and open many exciting opportunities for researchers in this field.