Morphological Changes in the Human Limbus with Pingueculae Imaged In-vivo with Ultrahigh Resolution Optical Coherence Tomography

Abstract

Purpose: : To evaluate the ability of ultrahigh resolution optical coherence tomography (UHR-OCT) to image non-invasively volumetric structural changes in the human limbus affected by pinguecula, such as lipid deposits, calcifications and elastotic degeneration of the conjunctival stromal collagen.

Methods: : A research-grade, UHROCT system was used to image in-vivo the human limbus with 3µm axial and 10µm lateral resolution at the rate of 47,000 A-scans/s. Volumetric tomograms of the limbus were obtained from control subjects without pinguecula, and experimental subjects with different grades of pinguecula. Images were processed with custom speckle denoising and segmentation algorithms, while commercial software (Amira, Visage Imaging) was used for the volumetric display of the imaging data.

Results: : Cross-sectional and volumetric images of the limbus and conjunctiva in the control subjects and at regions away from the pinguecula in experimental subjects revealed an intricate network of blood and lymph vessels, and clear delineation of the epithelial layer that thickens while transitioning from corneal to conjunctival epithelium. Images of the pinguecula showed different degrees of thickening and overall shape change of the limbus, with areas of high optical transparency (most likely lipid deposits), alternating with areas of irregular shape and high optical back-scatter (most likely calcium deposits and degenerated collagen).

Conclusions: : We demonstrated the capability of UHROCT to rapidly image in-vivo fine morphological features in human limbal and conjunctival tissue in healthy subjects and subjects with pinguecula. This technology has great potential for more thorough characterization of the structure and function of the normal and pathological human limbus.

Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • cornea: basic science • conjunctiva