Volumetry of the Ganglion-Cell Layer With a HD-OCT in Patients With Glaucoma and Correlation With Pattern Electroretinogram

Abstract

Purpose:: High-definition-OCT (HD-OCT) enables enhanced visualization of the retinal layers, including the retinal ganglion cell layer (RGCL) in most cases. We have designed an algorithm for automated segmentation of RGCL, based on macular HD-OCT scans. The function of retinal ganglion cells can be assessed with pattern electroretinogram (PERG). Glaucomatous eyes show a reduced RGCL as well as an impaired PERG. The aim was to analyse the thickness of macular RGCL in glaucoma and to compare it to electroretinographic signals and visual field defects.

Methods:: Three dimensional retinal imaging was performed using a second generation frequency domain HD-OCT system with a high axial resolution of 6 µm and up to 20kA-scans/second resulting in 128x512x1024 voxels per volume. 20 eyes of 20 patients with glaucomatous optic neuropathy and paracentral scotomas of different degrees were evaluated using 3D HR-OCT retinal imaging. Raster scanning of a 20 by 20 degree field and a depth of 2 mm was performed to provide a complete information about the macular ganglion cell layer. PERG was performed according to the International Society for Clinical Electrophysiology of Vision (ISCEV) standards for basic PERG with a stimulus field of 16 deg, a check size of 0.8 deg and a reversal rate of 2 Hz. The volumes of the RGCL (16 deg diameter) of the HD-OCT were correlated with PERG amplitudes and mean deviations of 10-2 Humphrey field analyser (HFA) visual fields (VF).

Results:: In a preliminary analysis of 8 eyes from 8 patients the correlation between the macular RGC volume and the PERG P50-N95 amplitude was 0.62 (r²). There was also a good correlation between the mean deviation of the VF and the PERG P50-N95 amplitude (r²=0.56) and the mean deviation of the VF and RGC volume (r²=0.53).

Conclusions:: We will present a correlation between the structure, assessed with rapid frequency-domain HR-OCT, and function, assessed with PERG and VF, of the macular retinal ganglion cell layer in glaucoma patients. Good correlations between RGCL volume and PERG as well as VF in a yet limited data set appear to validate the segmentation algorithm.

Keywords: ganglion cells • imaging/image analysis: clinical • electroretinography: clinical