Empirical Analysis Based on Light Attenuation Gradient of Wavelength Pairs for the Prediction of Skin Oxygen Status

Abstract

An empirical technique that allows noninvasive prediction of skin oxygen saturation (SO₂) using the absorption coefï¬cients of the preprocessing stored wavelength pairs is proposed. The highest probable SO₂ value is decided by selecting the bin containing wavelength pairs that produce the smallest variation in the distribution of the calculated attenuation gradient value. The performance this technique was evaluated using Monte Carlo simulated data. The simulation results revealed that this technique worked reasonably well even at low SO₂ condition with an overall mean error of not more than 2 %. This shows that the proposed analytic technique can potentially be used for measurement of the blood oxygen level of individuals with respiratory disease or with oxygen deprivation conditions such as hypoxic-hypoxia.