Thickness Mapping of Retinal Layers by Spectral-Domain Optical Coherence Tomography

Abstract

To report normal baseline thickness maps for 6 retinal layers generated by segmentation of spectral-domain optical coherence tomography (SD-OCT) images in normal subjects. Intersubject thickness variability and thickness variations in 9 macular sectors were established. Prospective cross-sectional study. SD-OCT imaging was performed in 15 normal subjects. Nineteen SD-OCT images were acquired, encompassing a 6 × 5-mm retinal area, centered on the fovea. Each image was analyzed using an automated segmentation algorithm to derive thickness profiles of 6 retinal layers. Thickness data obtained from all scans were combined to generate thickness maps of 6 retinal layers: nerve fiber layer, ganglion cell layer + inner plexiform layer, inner nuclear layer, outer plexiform layer, outer nuclear layer + photoreceptor inner segments, and photoreceptor outer segments. Mean and standard deviation of thickness measurements were calculated in 9 macular sectors and 6 retinal layers. Intersubject and intrasector thickness variations were established based on standard deviation of measurements. Minimum and maximum thickness of the nerve fiber layer were observed in the foveal and nasal perifoveal areas, respectively. The largest thickness variation among subjects and intrasector variability were observed in perifoveal areas. Thickness of the ganglion cell layer + inner plexiform layer and intersubject thickness variability were largest in parafoveal areas. The inner nuclear layer thickness was relatively constant in parafoveal and perifoveal areas and intrasector thickness variations were largest in the foveal area. The outer plexiform layer thickness was relatively constant in foveal and parafoveal areas and higher than in perifoveal areas. Intersubject thickness variability in inner nuclear layer and outer plexiform layer was relatively uniform in all macular sectors. The outer nuclear layer + photoreceptor inner segments thickness map displayed maximum thickness in the foveal area and intersubject thickness variability was largest superior to the fovea. Thickness of the photoreceptor outer segments layer, thickness variations among subjects, and intrasector thickness variability were relatively constant. There was a significant correlation between total retinal thickness derived by thickness mapping and SD-OCT commercial software. Normal thickness maps for 6 retinal layers were generated and thickness variations among subjects and macular areas were assessed. This technique is promising for investigating thickness changes attributable to disease in specific retinal layers and macular areas.