The Photoplethysmographic Biosensors can be used to monitor various physiological parameters like arterial oxygen saturation, glucose level in the blood, heart rate, blood pressure etc. But the Photoplethysmographic signals could be contaminated by various noise sources. The performance limitation due to the motion artifacts is very high compared to that due to other noise sources. Conventional filtering techniques are incapable to get rid of motion artifacts effectively and completely due to the frequency overlapping between the motion artifacts and the clean Photoplethysmographic signal. This work focuses on the performance enhancement of the Photoplethysmographic Biosensor by removing the effect of motion artifacts caused by the voluntary movements of the individual. The obj ective was achieved by developing an algorithm which is capable of providing motion artifact-free Photoplethysmographic signal during various physical activities of an individual without compromising any signal characteristics. We have developed three motion artifact removal algorithms: Butterworth-Wavelet Algorithm, Adaptive Normalized Least Mean Square Algorithm and Adaptive Recursive Least Square Algorithm. The three algorithms were performed on hundred samples collected from twenty-five subj ects under four fingertip movements like shivering, vertical movement, horizontal movement and applying pressure. Based on the SNR analysis, Butterworth-Wavelet Algorithm was found better in providing high SNR
A. Luke, Shereena Shaji, and Menon, U., “Performance Enhancement of a Photoplethysmographic Biosensor Using Efficient Signal Processing Techniques”, in 2018 3rd International Conference for Convergence in Technology (I2CT), 2018.