The effect of basic assumptions on the tissue oxygen saturation value of near infrared spectroscopy

Tissue oxygen saturation (StO(2)), a potentially important parameter in clinical practice, can be measured by near infrared spectroscopy (NIRS). Various devices use the multi-distance approach based on the diffusion approximation of the radiative transport equation [1, 2]. When determining the absorption coefficient (μ (a)) by the slope over multiple distances a common assumption is to neglect μ (a) in the diffusion constant, or to assume the scattering coefficient (μ(s)) to be constant over the wavelength. Also the water influence can be modeled by simply subtracting a water term from the absorption. This gives five approaches A1-A5. The aim was to test how these different methods influence the StO(2) values. One data set of 30 newborn infants measured on the head and another of eight adults measured on the nondominant forearm were analyzed. The calculated average StO(2) values measured on the head were (mean ± SD): A1: 79.99 ± 4.47%, A2: 81.44 ± 4.08%, A3: 84.77 ± 4.87%, A4: 85.69 ± 4.38%, and A5: 72.85 ± 4.81%. The StO(2) values for the adult forearms are: A1: 58.14 ± 5.69%, A2: 73.85 ± 4.77%, A3: 58.99 ± 5.67%, A4: 74.21 ± 4.76%, and A5: 63.49 ± 5.11%. Our results indicate that StO(2) depends strongly on the assumptions. Since StO(2) is an absolute value, comparability between different studies is reduced if the assumptions of the algorithms are not published.