Analysis and quantification of collagen organization with the structure tensor in second harmonic microscopy images of ocular tissues

The important biological role of collagen-based tissues and the changes produced in the fiber distribution under particular situations (surgery, pathology, external damage, etc.) require tools for the analysis of the collagen organization that might potentially help in early diagnoses. Since collagen structures provide efficient second harmonic generation (SHG) signals, SHG microscopy has emerged as a powerful technique to visualize collagen fibers and qualitatively discriminate normal from abnormal tissues. Here we propose a quantitative method based on the structure tensor to quantify the different organization of collagen patterns in SHG images of ocular tissues. Results show that well-organized collagen distributions present a high degree of isotropy (DoI), a dominant orientation (PO), and a low structural dispersion (SD). On the other hand, the PO vanishes when the collagen tissue is not organized as a consequence of an increase in the SD and a decrease in the DoI. The proposed method is also able to discriminate partially organized samples. The combination of SHG microscopy and the structure tensor is a useful method to objectively classify collagen distributions. Clinical applications of this technique could help in the diagnosis and tracking of pathologies related to collagen disorders in connective tissue.