Abstract
Purpose :
Since the particular distribution of collagen fibers within the stroma is related to corneal transparency, any alteration might seriously compromise the optical properties of the cornea. The use of second harmonic generation (SHG) microscopy demonstrated that healthy and pathological collagen-based tissues present different organization. We propose to analyze corneal tissues with SHG microscopy to quantitatively discriminate the collagen fibers organization in pathological from healthy ones.
Methods :
A custom multiphoton microscope was used for SHG imaging of ex-vivo corneal tissues from human donors. SHG images from corneas affected by keratoconus, edema and bullous keratopathy were recorded. These were compared to SHG images of healthy corneas used as control. The structure tensor was used to characterize the collagen fiber dispersion information from SHG images. Different quantitative information parameters were obtained from each sample: preferential orientation (PO), structural dispersion (SD) and the degree of isotropy (DoI) of the collagen fibers (lamellae).
Results :
Control corneas showed systematically a regular lamellar distribution with high DoI and a low SD. The PO presented a Gaussian distribution with the maximum corresponding to the dominant direction of the fibers. However, this pattern turned into a non-organized arrangement in all tested pathological corneas. The DoI was lower and consequently the SD noticeable increased. Moreover, there was not a clear preferential orientation.
Conclusions :
SHG microscopy was used to image collagen architecture in corneal tissues under different experimental conditions. From the images, the structure tensor approach provided information on the fiber spatial organization in both pathologic and normal corneal tissues based on several quantitative parameters. This permits to differentiate between normal and pathological corneas, as well as to establish discrimination numerical scales. Although, currently SHG can only be applied “in vitro”, future implementations of this techniques in clinical environments might represent a better tool for corneal disease diagnoses.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.