1.1 Tissue Structure and Scattering Properties
Soft tissue is composed of closely packed groups of cells entrapped in a network of fibers through which interstitial liquid percolates. At a microscopic scale, the tissue components have no pronounced boundaries, thus tissue can be considered as a continuous structure with spatial variations in the refractive index. To model such a complicated structure as a collection of particles, it is necessary to resort to a statistical approach. The tissue components that contribute most to the local refractive index variations are the connective tissue fibers (either collagen or elastin forming, or reticulin forming) that form part of the noncellular tissue matrix around and among cells, and cell membrane; cytoplasmic organelles (mitochondria, lysosomes, and peroxisomes); cell nuclei; and melanin granules. Figure 1 shows a hypothetical index profile formed by measuring the refractive index along a line in an arbitrary direction through a volume of tissue and corresponding to the statistical mean index profile. The widths of the peaks in the actual index profile are proportional to the diameters of the elements, and their heights depend on the refractive index of each element relative to that of its surroundings. This is the origin of the tissue-discrete particle model. In accordance with this model, the index variations may be represented by a statistically equivalent volume of discrete particles having the same index, but different sizes.
The refractive indices of tissue structure elements, such as the fibrils, the interstitial medium, nuclei, cytoplasm, organelles, and the tissue itself can be derived using the law of Gladstone and Dale, which states that the resulting value represents an average of the refractive indices of the components related to their volume fractions: where ni and fi are the refractive index and volume fraction of the individual components, respectively, and N is the number of components.
The statistical mean index profile in Fig. 1 illustrates the nature of the approximation implied by this model.