Quantification of collagen fibril alignment and density in second harmonic generation images of cellular collagen matrices

Collagen is one of the most abundant proteins found in the body. Its quantification during cellular induced remodeling is important in differentiating normal and diseased tissue development. Collagen can be used for specific applications related to disease diagnostics as well as tissue engineering. Within tissue, cells proliferate and remodel their surrounding extracellular matrix changing their environment including collagen. Collagen is not only biocompatible but also strong due to its unique structure. Since the ECM is a complex system of collagen interacting with cells, a dynamic technique is needed to quantify the biodegradation and deposition of collagen. In this project, we present the use of a combination of non-linear optical imaging and image processing as a non-invasive technique for quantifying the collagen fibril density and alignment within a model extracellular matrix.; To model an extracellular matrix, collagen gels were created and examined alone or when seeded with fibroblast cells. Two-photon excitation was used to simultaneously image both the cells and the collagen. Second Harmonic Generation (SHG) was used to image collagen type I fibrils and to monitor collagen trafficking. Endogenous two-photon excited fluorescence (TPEF) was detected from the cells.; In order to quantify the dynamic changes in the density of collagen fibrils within the acquired images, the Power Spectral Density (PSD) based on the Fourier transform was used. Furthermore, an Orientation Index (OI) was calculated from the PSD to assess the percentage of aligned fibrils along a dominant direction. Changes in the spatial frequency-dependent PSD at specific angles, provide insight into the changes in fibril size during collagen reorganization. Local fibril alignment was determined by segmenting an image into tiles and performing the Hough Transform to find the level and direction of alignment on each tile. The amplitudes and directions from all the tiles were summed to give a distribution of local alignment and the entropy was calculated as a measure of organization.; We found that collagen reorganization was highly dependent on the density of cells present. Specifically, in the collagen gels seeded with a high density of fibroblasts, we observed initially significant levels of collagen degradation, followed by deposition of highly aligned fibers. In the collagen matrices seeded with a lower density of cells the differences in collagen morphology are not as distinct, exhibiting only an overall increase in fibril density, as assessed by the PSD-based Orientation Index. Entropy calculations based on the Hough Transform analysis of localized image regions indicate that the levels of alignment are particularly enhanced in areas immediately surrounding the cells.; Development of non-invasive techniques for the characterization of collagen fibril density and alignment is a step forward in understanding cell-matrix interactions in the context of disease diagnostic and tissue engineering applications.