Abstract
Purpose :
Optical coherence tomography (OCT) measurements of retinal thickness (RT) and choroidal thickness (RT) play crucial roles in the diagnosis, monitoring, and management of retinal and choroidal diseases. It is important to evaluate RT and CT from the new swept source OCT (SS-OCT) and that from spectral domain OCT (SD-OCT) to allow accurate comparison of measurements. The primary objective was to compare RT and CT obtained from SS-OCT and SD-OCT and to describe the factors that influenced the variation. The secondary objective was to compare RT and CT maps produced by different scanning protocols of the SS-OCT.
Methods :
In a prospective cohort study of 100 normal participants, OCT scans were performed sequentially with the Topcon DRI OCT-1 and Spectralis OCT using standardized imaging protocols. RT and CT in the Early Treatment Diabetic Retinopathy Study (ETDRS) subfields were obtained from the autosegmentation software and independently measured by masked reading center-certified graders respectively. Paired t-tests and intraclass correlation coefficients (ICCs) were used to compare the measurements.
Results :
The mean central RT differed significantly between DRI OCT-1 and Spectralis (DRI OCT-1: 240.0µm; Spectralis: 273.1 µm; p<0.001), with a mean difference of 33.1µm. The inter-device ICC for RT was 0.957. The mean central CT also differed significantly between DRI OCT-1 and Spectralis (DRI OCT-1: 265.6µm; Spectralis: 313.7µm; p<0.001), with a mean difference of 48.1µm. The inter-device ICC for CT was 0.757. Similar results were observed in all ETDRS subfields. After manual adjustment of segmentation boundaries for central subfield in the DRI OCT-1, the mean central CT for DRI OCT-1 increased from 265.6µm to 295.4µm (p<0.001), and the difference between DRI OCT-1 and Spectralis decreased from 48.1µm to 18.3µm (decrease of 62.0%; p<0.001). In all sectors, RT and CT produced by the three-dimensional (3D) and radial scanning protocols of the SS-OCT were similar (mean differences: 0.6-3.6µm for RT; and generally under 10µm for CT; both p<0.001).
Conclusions :
RT and CT measurements obtained from SS- and SD-OCT are different, and should be accounted for when comparing measurements. For CT, the difference can be reduced, but not eliminated, by manual adjustment of segmentation boundaries by trained graders.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.