En Face Optical Coherence Tomographic Imaging of the Optic Nerve Head: In Vivo Comparison with Scanning Laser Ophthalmoscopy.

Abstract

Abstract: : Purpose: En–Face Optical Coherence Tomography (efOCT) is a novel and promising technology that provides high–resolution three–dimensional images of retinal structures. We compared stereometric parameters of the optic nerve head (ONH) as measured by efOCT and confocal Scanning Laser Ophthalmoscopy (cSLO). Methods:A prototype efOCT, developed by the Institute of Ophthalmology and the Applied Optical Group at the University of Kent, was used in a pilot study to scan four eyes from four normal subjects. The efOCT comprises an interferometer and superluminescent diode at a wavelength of 850 nm and a power to the eye of 500 µW. The instrument has two channels: confocal and OCT. Confocal and efOCT (C–scan/transversal) images are collected simultaneously. The efOCT scanning method is different to that used in conventional longitudinal Optical Coherence Tomography (OCT) and similar to that used in cSLO (producing C–scan image stacks). Successive C–scans at different depths were performed and a stack of images was generated. efOCT C–scans of the ONH were used to produce composite three–dimensional images, which were quantified and analysed using software specifically designed to incorporate a reference plane and a contour line akin to Heidelberg Retina Tomograph (HRT II). Eyes were also scanned using the commercially available HRT II. Two experienced observers drew contour lines in the images derived from each device. In the efOCT, the contour line was placed over the image produced by the confocal channel. Disc area (DA), cup area (CA), rim area (RA) and CA/DA ratio were calculated for both devices. Measurements were compared by means of correlation coefficients and Bland & Altman plots. Results:Correlation between measurements was R²=0.8 (P=0.08) for DA, R²=0.9 (P=0.04) for CA, R²=0.9 (P=0.04) for RA, and R²=0.9 (P=0.03) for CA/DA ratio. The mean difference and standard deviation (±) of differences between methods were 0.1 ± 0.1mm² for DA, 0.03 ± 0.08mm² for CA, 0.08 ± 0.07mm² for RA, and 0.003 ± 0.03 for CA/DA ratio. Conclusions:Very good agreement was found between the analyses of efOCT and HRT II images. The dual presentation of confocal and OCT images, and its high depth resolution, gives great potential for detailed three–dimensional analysis of retinal structures.

Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)