Purchase this article with an account.
A.C. Walsh, P.G. Updike, J.F. Updike, S.R. Sadda; True Retinal Thickness with 3D–OCT . Invest. Ophthalmol. Vis. Sci. 2006;47(13):4031.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent advances in OCT technology have demonstrated that conventional OCT software often incorrectly identifies the inner segment/outer segment junction as the outer retinal boundary. Because of this, clinical OCT machines typically underestimate the true retinal thickness. Fourier domain (FD) OCT, on the other hand, has better axial resolution (<6 microns), operates at more than 60 times the capture rate, and can acquire a dense, three–dimensional grid of data in less time than conventional OCT. These advantages may improve the ability of FD OCT to detect the true outer retinal boundary and measure actual retinal thickness. Proving this, however, is complicated by the knowledge that conventional OCT represents a flawed standard for comparison. Therefore, the aim of this study is to compare the performance of a new FD OCT device and software algorithm to a new gold standard for outer retinal boundary delineation – manual measurement by a trained reading center grader.
3D–OCT data were collected from 20 eyes of 10 normal subjects using an investigational FD OCT device integrated into a nonmydriatic camera. Reading center graders used custom software to outline the inner and outer retinal boundaries in all 3D–OCT scans with a computer mouse. Two graders evaluated all scans independently and their results were compared with kappa coefficients to determine intergrader variability. Retinal boundaries from their evaluations (> 1 million data points) were used as the gold standard for comparison to results from an automated 3D–OCT analysis system called OCTANE.
Intergrader agreement for reading center assessments of all points in the 3D–OCT scans was excellent with an average difference between graders of less than 6 microns (the resolution of the FD OCT machine). OCTANE retinal thickness measurements correlated well with reading center measurements (p<.001). On average, the position of the inner and outer retinal boundaries determined by OCTANE deviated less than 10 microns from the reading center standard.
OCTANE analysis of 3D–OCT data provides an accurate measure of retinal thickness in normal patients when compared to a reading center gold standard. The increased speed and segmentation accuracy of 3D–OCT with OCTANE may be useful in the detection of retinal thickening in patients with retinal diseases.
This PDF is available to Subscribers Only