Prediction of Calibration Parameters of the Oxygen Saturation Estimation Equation by Optical Recording on Smartphones.

Abstract

This paper demostrate the calculate for calibration parameters of the Percent Oxygen Saturation (%¿¿¿2 ) estimation equation by means of a multiple linear regression type supervised machine learning algorithm with the objective of being able to use a smartphone as an oximeter, for this a training and test data set composed of predictor variables representing the characteristics of the smartphones used in this development is used, in addition to the statistical parameters extracted from two thousand videos, predicting values of %¿¿¿2 with an error of less than 2% with respect to a standard oximeter.