A data reduction scheme for improving the accuracy of oxygen saturation calculations from spectrometric in vivo measurements.

A developing tool in ophthalmological research and practice is the study of the oxygen saturation in retinal vessels because this measurement contains important diagnostic information about the metabolism at the human ocular fundus. Imaging spectrometry enables locally resolved measurements of the oxygen saturation simultaneously in arteries and veins. The vessel reflex, imaging errors of the detection system and the noise of the signal result in substantial uncertainties in the calculated data. We present a data processing method for the improvement of the primary information leading to a reduction of the uncertainties of the derived oxygen saturation.