Reduction of motion artifacts from pulse oximeter signals using tunable Q-factor wavelet transform technique

This is the authors' abstract. We don't add key points for this paper.

Pulse Oximeter (PO) employed in critical care units is crucial equipment to measure the vital parameters like oxygen blood saturation levels and heart rate of the patient. Using PO device, required medical data is acquired using the photoplethysmographic (PPG) data utilizing PPG sensors attached on forehead/ to finger/at earlobe of the patient and then Ratio parameter (R) is computed pertaining to amplitudes of acquired red and IR PPG signals. Further, `R' is used to estimate oxygen saturation levels with the help of calibration curve. Subject movements while recording the medical data may result in erroneous estimation of required estimation parameter and in turn may result in wrong diagnosis by the clinician. Reduction of Motion Artifacts (MA) component from raw PPG data recorded may guarantee error-free measurement of oxygen blood saturation level (SpO 2 ). MA's can be removed from raw PPG signal (corrupted) using band pass filtering method, but the persisting in-band noise component cannot be removed. In this paper, authors propose a filtering method using tunable Q-factor wavelet transform (TQWT) to remove MA components. Advantage of TQWT sytems from the fact that, the realization of practical narrow band pass filter with a specific Q-factor value can be designed, which motivated the authors to use for this application. Experimental results have shown a good acceptance for the proposed method as the MA reduced PPG signals obtained are having efficient morphological features. SpO 2 is estimated from MA reduced PPGs by utilizing the calibration curve. The superiority of proposed technique has been proved by comparing the experimental results with results obtained using basic least mean squares (LMS) method. Signal data can be acquired with different MA components (bending, horizontal and vertical movements of patient's finger) is considered for experiment analysis. Obtained SpO 2 parameter calculations proved the efficacy of estimation technique in measurement of reliable and accurate SpO 2 , helpful for medical diagnosis.

May enhance motion-robust SpO2 monitoring in wearables; leaves open prospective clinical validation before practice change.