Computational analysis of light scattering from collagen fiber networks

Neoplastic progression in epithelial tissues is accompanied by structural and morphological changes in the stromal collagen matrix. We used the Finite-Difference Time-Domain (FDTD) method, a popular computational technique for full-vector solution of complex problems in electromagnetics, to establish a relationship between structural properties of collagen fiber networks and light scattering, and to analyze how neoplastic changes alter stromal scattering properties. To create realistic collagen network models, we acquired optical sections from the stroma of fresh normal and neoplastic oral cavity biopsies using fluorescence confocal microscopy. These optical sections were then processed to construct three-dimensional collagen networks of different sizes as FDTD model input. Image analysis revealed that volume fraction of collagen fibers in the stroma decreases with neoplastic progression, and statistical texture features computed suggest that fibers tend to be more disconnected in neoplastic stroma. The FDTD modeling results showed that neoplastic fiber networks have smaller scattering cross-sections compared to normal networks of the same size, whereas high-angle scattering probabilities tend to be higher for neoplastic networks. Characterization of stromal scattering is expected to provide a basis to better interpret spectroscopic optical signals and to develop more reliable computational models to describe photon propagation in epithelial tissues.