Intraobserver and Interobserver Reproducibility in the Evaluation of Optic Disc and Retinal Nerve Fiber Layer by Rtvue Fourier-Domain Optovue

Abstract

Purpose: : To assess intraobserver and interobserver reproducibility of morphometric measurements on optic disc and retinal nerve fiber layer in glaucoma (POAG), ocular hypertensive(OHT) and normal subjects by means of RTVue Fourier-Domain OCT. Study

Results: : ICC showed worst scores on rim (0.02 to 0.48) and nerve head (-0.01 to 0.17)volume in all of the calculations. Best scores were found on RNFL related parameters (0.94 to 1),either on intraobserver-interimage or interobserver-intra/interimage evaluations. The CL drawing system did not affect reproducibility of these parameters. The parameters cup-disc horizontal and vertical ratio showed substantial to very good correlation (0.67 to 0.88) both in intra- and interoperator evaluation, except when the 3D baseline CL was used. In general, in all of the evaluations, the 3D baseline CL drawing had worse scores than OCT and video-baseline drawing systems. Interimage (33.3 to 47,2% parameters ICC > 0.7) showed a higher variability than intraimage (50% parameters ICC > 0.7) evaluation.

Conclusions: : Measurement of RNFL parameters by means of RTVue is highly reproducible and slightly operator-dependent. Optic nerve head measurements are less reproducible and more operator-dependent than RNFL values. 3D baseline showed higher variability than OCT and video-baseline. Since the intra-image comparison gives the best ICC scores, image acquisition is more a source of variability than contour line drawing.

Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • optic disc • imaging/image analysis: clinical