Abstract
Purpose: :
To evaluate 1) the accuracy of machine-measured central foveal thickness (CFT) compared to manual measurements and 2) changes in CFT and photoreceptor layer (PRL) thickness in eyes with various macular diseases compared to normal eyes using a spectral-domain optical coherence tomography (SD-OCT).
Methods: :
Data from 89 eyes of 89 patients who underwent imaging during June of 2009 with a SD-OCT (Spectralis) with diagnoses of epiretinal membrane (ERM, n=28), dry age-related macular degeneration (AMD, n=20), diabetic macular edema (DME, n=10), chronic central serous chorioretinopathy (CSC, n=11) and normal (controls, n=20) were collected. For each eye, SD-OCT images were opened using the ImageJ program. Within ImageJ, the CFT and the height of PRL at the fovea were manually measured using longitudinal reflectivity profiles. Automated CFT was compared with manual CFT. Additionally, the mean height of PRL in various macular diseases was compared with mean value of controls to assess how PRL thickness may be affected in macular diseases. The height of PRL was measured as the distance from the start of the outer nuclear layer to the top of retinal pigment epithelium at the fovea.
Results: :
For all patients, mean automated CFT was 302 µm and manual CFT was 296 µm (p=0.96). The mean automated and manual CFT for each diagnosis did not differ statistically and were as follows: Control-233 µm and 230 µm (p=0.56); ERM-368 µm and 378 µm (p=0.05); dry AMD-234 µm and 224 µm (p=0.27); DME-397µm and 394µm (p=0.90), and CSC-268 µm and 260µm (p=0.16). In eyes with ERM and DME, the mean manual CFTs were greater than that of controls (p=0.00000 and p=0.00001, respectively). Compared to mean PRL thickness in control eyes (164 µm), the mean PRL thickness in ERM (248 µm) and DME (323 µm) were greater (p=0.00001 and p=0.00000, respectively) while dry AMD (146 µm) was thinner (p=0.017).
Conclusions: :
The automated CFT by a SD-OCT correlated well with manual measurements in normal eyes and in eyes with types of macular diseases included in this study. In ERM and DME, the mean CFT as well as PRL were thicker than normal eyes. In dry AMD, the mean CFT did not differ significantly from normal eyes but the height of PRL was thinner. Therefore, various macular diseases may affect not only the CFT but also PRL differently.
Keywords: retina • imaging/image analysis: clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)