Purchase this article with an account.
R. P. Danis, A. Domalpally, A. Kulkarni, Q. Peng; Correlation of Visual Acuity to Outer Retinal Thickness Measurements in Spectral Domain OCT Images of Various Retinal Diseases. Invest. Ophthalmol. Vis. Sci. 2010;51(13):277.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess the relationship between visual acuity (VA) and outer retinal thickness parameters evaluated from spectral domain optical coherence tomography (SDOCT) images in various retinal diseases in a prospective clinical care cohort.
136 eyes from 73 patients were scanned using the Topcon 3D OCT-1000 using the 512X128 scan protocol for standard of care retinal disease management and had visual acuity measured per clinical routine with a Snellen chart (and converted to LogMAR for analysis). Using the manufacturer’s display application, photoreceptor length (PRL) was measured from the B scan passing horizontally through the foveal center using caliper measurements from the outer plexiform layer to the retinal pigment epithelium (RPE) at three points: centerpoint and the nasal and temporal limits of the central subfield (CSF). Ungradable points were excluded. All scans were evaluated twice by two examiners and any disagreements were reviewed by a third examiner. Automated measurements were obtained using the manufacturer’s software for CSF retinal thickness and photoreceptor outer segment length (OSL) - defined as the distance from the junction of photoreceptor inner segment outer segment (IS-OS) and the RPE.
Mean PRL was available for 119 (87%) of scans. The intraclass correlation (ICC) between the two examiners for both the mean PRL was 0.71. The Pearson’s correlation coefficient for logMAR VA and mean PRL was -0.56, PRL at centerpoint was -0.51, for OSL -0.43, and for CSF thickness the Pearson’s was 0.2. The correlation varied greatly across disease categories (range -0.28 to -0.67 for PRL and VA) but interpretation was difficult due to small sample size in each category.
PRL and OSL had a better correlation with VA than did CSF retinal thickness using a manual measurement method. A large proportion of images may not be gradable for these features which may limit the usefulness of these variables in clinical trials.
This PDF is available to Subscribers Only