ABSTRACT
Human tissue oxygen saturation (rSO2) can be monitored non-invasively and in real time by near infrared spectroscopy (NIRS) based on spatially-resolved spectroscopy. To expand the clinical applications of the NIRS oximeter developed by our group based on the above principle, the accuracy of rSO2 must be ensured and enhanced as far as possible. In the present paper, the influences of the probe configuration, especially the distance between the detectors and the wavelength discreteness of the light source, on the accuracy of rSO2 were discussed. The results indicate that (1) to obtain rSOz accurately, two detectors need to be used, where the distance between them should be in the range of 5-20 mm and they should be both at least 20 mm apart from the light source; (2) there can be significant error in rSO2 (> 10%) induced by the discreteness of the two emission wavelengths especially the shorter one of the light source, so the real emission wavelengths must be accurately measured and the corresponding extinction coefficients of deoxygenated and oxygenated hemoglobin (Hb and HbO2) must be used in order to avoid this error. The above conclusions can be the guidance to optimally design the probe, which has been achieved in our NIRS oximeter.
