Reduction of Pigmentation Dependent Errors using Spectroscopic Pulse Oximetery

PROJECT SUMMARY/ ABSTRACT Pulse oximetry is a powerful clinical tool for clinical evaluation of patients with cardiopulmonary compromise, however, multiple clinical studies have indicated pulse oximetry suffers from overestimation errors in darkly pigmented individuals, which is a critical mistake in screening for a cardiopulmonary disease, including COVID- 19, and could contribute to reports of outcome disparities in ethnic minorities. Aimloxy, LLC has been developing proprietary hardware and firmware for High Frequency Pulse Sensing Technology (HF-PST), which represents a next-generation evolution of photoplethysmography (PPG) pulse sensing technologies, which have conventionally served as the backbone of pulse oximetry. The Optical Diagnostic Research Laboratory (ODRL) at Temple University is involved in the development of optical sensing technologies for low-resource settings and has developed mobile-phone based technologies that include strategies for reducing pigmentation dependent inaccuracy in transcutaneous sensing of bilirubin levels in African-origin infants. The ODRL, Aimloxy, along with collaborators from Temple University Hospital’s Department of Emergency Medicine and Fox School of Business’ Center for Data Analytics will team together to perform initial development and benchtop characterization of a next-generation high frequency spectroscopic pulse oximeter capable of reducing pigmentation dependent errors in PPG and oximetry signals, conduct an initial human subjects feasibility study on a multi-ethnic population, and incorporate refined machine learning algorithms to improve measurement accuracy in pigmented individuals. The successful completion of this phase I proposal will set the stage for a full-scale study in phase II to develop a commercial prototype and perform multicenter trials to characterize accuracy and compare performance against traditional techniques in darkly pigmented individuals, ultimately helping reduce healthcare disparities in racial minorities.PROJECT NARRATIVE In this proposal, we seek to develop a next-generation high frequency spectroscopic pulse oximeter capable of reducing pigmentation dependent errors in PPG and oximetry signals, and then further embed Aimloxy LLC proprietary technologies that facilitate improved machine learning prediction of hypoxia. Ultimately, the successful completion of this work will set the stage for clinical validation of a simple finger clip style product with a next generation hardware design ideally suited to inform powerful machine learning algorithms, which can ultimately reduce healthcare disparities.