Mapping Blood Oxygen Saturation using a Multi-Spectral Imaging System

Recent advances in imaging spectroscopy provide the opportunity for mapping the oxygen saturation of blood in skin with high accuracy, large spatial coverage, small spatial resolution, and high update rate. A four—wavelength algorithm, specifically designed to compute the oxygen saturation of hemoglobin, in vivo, from a set of narrow-band visible images was used to analyze various skin tissue disorders. To illustrate the spatial capability of this algorithm, mapping of the oxygen saturation of normal skin, hypoxic tissue and various skin lesions was performed using reflectance spectroscopy, demonstrating the spatial resolution of the images of blood oxygen in the tissues. To explore the accuracy of the algorithm, Monte—Carlo modeling was used to generate reflectivities of skin with known parameters. These reflectivities were used to evaluate the limiting effects of quantization error, photon noise, and finite filter bandwidth on the accuracy of the algorithm. In addition, a signal—to—noise (SNR) analysis was performed to determine the illumination requirements. It is shown that accurate maps of blood oxygen can be produced with good spatial resolution.