Lookup-table-based inverse model for mapping oxygen concentration of cutaneous microvessels using hyperspectral imaging.

Hyperspectral imaging combining with skin optical clearing technique provides a possible way to non-invasively monitor hemodynamics of cutaneous microvessels. In order to estimate microvascular blood oxygen saturation, in this work, a lookup-table-based inverse model was developed to extract the microvascular optical and physiological properties using hyperspectral analysis. This approach showed a higher fitting degree than currently existing hyperspectral analysis methods (i.e. multiple linear regression and non-negative least square fit) in estimating blood oxygen saturation. Hypoxic stimulation experiment showed that calculated results were in accordance with physiological changes, and the relative changes of estimated oxygen saturation indicated this method appeared to be more sensitive to blood oxygen fluctuation. And a simulated blood model was used for verification here, indicating this method also showed a good accuracy in determining oxygen saturation from the simulated spectra.