Quantification of Collagen Fiber Orientation and Organization by Singular Value Decomposition

Abstract

Second-harmonic generation microscopy is a new imaging tool that allows analyzing non-centrosymmetric structures, such as collagen. This method can be used to detect structural organization of human ovarian extracellular matrix. Collagen remodeling can be used as early indicator of disease states, such as tumor development and invasiveness. Scoring structural characteristics like main collagen orientation and organization do allow differentiating between healthy and cancerous tissue. In this report, it is proposed a new method to calculate the main orientation angle as well as the degree of collagen fibers organization. The algorithm is based on the analysis of image Fourier transform modulus, characterizing spectral distribution by using singular value decomposition. The preliminary results obtained in this work demonstrate that the method allows to discriminate healthy from cancerous tissue.