Noninvasive determination of blood oxygenation in the brain by near-infrared reflectance spectroscopy
Hielscher, Andreas Helmut
Tittel, Frank K.
Doctor of Philosophy
Continuous monitoring of blood oxygenation in brain is of great clinical interest. Hyperoxia and hypoxia, even over a short period of time, can lead to irreversible neurological damage. The recent progress in techniques to spectroscopically characterize turbid media has led to new possibilities to determine blood oxygenation by measuring the absorption coefficient of blood-containing biological tissues. In this work the influence of the air-tissue interface on the accuracy of the determination of tissue optical properties is examined. Different theoretical expression derived from diffusion theory under three commonly assumed boundary conditions are compared to each other, to Monte Carlo simulations and to experiments on tissue phantoms. It is shown that a simple fitting algorithm, which is based on the time-resolved diffusion theory with the zero boundary condition, can be used to determine the absorption coefficient of homogeneous tissues with high accuracy. When this algorithm is applied to heterogeneous tissues, the determined absorption coefficient has to be interpreted as an apparent absorption coefficient. For the case of layered tissues it is demonstrated, through simulations and experiments, that in most practical situations the apparent absorption coefficient equals the absorption coefficient of the underlying medium. For example, the skin, the skull and the meninges do not affect the determination of the absorption properties of the deeper located brain tissue. Furthermore, it is established that in the case of tissues with embedded, high-absorbing, blood vessels, the apparent absorption coefficient can be approximated by a volume-weighted sum of the blood absorption and the tissue-background absorption. This finding leads to a correction of commonly applied blood-oxygenation-determination algorithms, which neglect the background absorption. Experiments on tissue phantoms and biological tissues in vivo exemplify the use of the improved algorithm. Finally, it is illustrated how photon density waves can be used to distinguish between life-threatening hemorrhages and less dangerous blood-oxygenation decreases in parts of the brain. An analytical solution for the wave propagation in the present of a spherical optical inhomogeneity is derived and used to estimate the sensitivity and specificity of photon-density-wave based tomography.
Biomedical engineering; Biophysics