Determining the optical properties of the human brain using a two-layer diffusion model for non-invasive, frequency domain optical measurements

The concentrations of oxyhemoglobin [HbO] and deoxyhemoglobin [Hb] in tissue and their temporal dynamics are an important indicator of cognitive function in the brain. Near-infrared (NIR) techniques have been employed to extract the concentration levels of these tissue chromophores due to their non-invasive feature. An integral part of this extraction procedure is a model of light propagation in turbid medium such as human tissue. In this study, we apply the diffusion model since the nature of light-tissue interaction for tissues probed with near-infrared light is highly scatter-dominated. In such a model of light propagation, the quantities of interest are the absorption μ a and the scattering coefficient μs. From the absorption coefficients, the chromophore concentrations can be determined through their molar extinction coefficients. However, the model assumes a homogeneous issue structure and inevitably presents a limitation in the extraction process.;In this study, a two-layer model of the head is proposed to account for the heterogeneity in the tissue architecture of the head in recovering the absorption and reduced scattering coefficients (optical properties) of the brain. The model assumes a superficial layer of finite thickness (first layer) situated over a semi-infinite homogeneous structure (second layer). Applied to the head, the scalp and the skin are lumped together in the first layer and the brain comprises the latter. Values of intensity amplitude (AC) and phase (&thetas;) are acquired in the frequency domain at multiple source-detector distances in a reflectance geometry. The optical properties are then recovered by performing a nonlinear regression to Monte Carlo-generated data for the AC and phase of the detected light. It is shown that the recovered optical properties using this two-layer model exhibits better agreement with the simulations than those recovered from a homogeneous model.;The study hopes to provide an added avenue of refinement to a recent protocol that was developed to perform absolute measurements of cerebral hemoglobin concentration.