Abstract
Purpose::
. To study the agreement between optical coherence tomography (StratusOCT) and digital stereophotographs in the assessment of the vertical cup-to-disc ratio (VCDR) and investigate how the cup-offset reference plane should be changed to improve such an agreement.
Methods::
. StratusOCT was used to image the optic disc by the Fast Optic Nerve Head (ONH) acquisition protocol; the VCDR was measured after manual correction of the ONH edges. Thirty-degree digital stereophotographs were taken by a Topcon camera and analyzed by Imagenet 2000 using the stereo tool; two masked examiners analyzed the coupled images and measured the VCDR by an external application. The mean of the their estimation was used as a final grading. The two methods were compared by Bland-Altman plots with 95% limits of agreement (LAE). Subsequently, for each eye, the standard cup-offset (= 150 microns above the retinal pigment epithelium [RPE]) was changed so that the VCDR measurement would correspond to that obtained by stereophotograph.
Results::
. Fifty-five eyes (55 patients) were studied; the mean VCDR was 0.48±0.19 when assessed by stereophotographs and 0.52 ±0.2 when assessed by StratusOCT (p = 0.0005). In 36 cases (65.4%) such a difference was less than 0.1, a value that can be considered clinically not relevant. In 15 cases (27.2%) StratusOCT overestimated the VCDR by more than 0.1 and in 4 cases (7.2%) it underestimated the VCDR by more than 0.1. The 95% LAE were -0.21 +0.13. An ideal agreement was found by changing the StratusOCT default cup-offset to values ranging from -100 to 305 microns, depending on the cup morphology. In eyes with a deep cup (i.e. deeper than the RPE reference plane) the same VCDR as measured by stereophotographs would have been achieved by moving downward the cup-offset to a mean value of 69.1±98.1 microns, while in eyes with a shallow cup (i.e. more superficial than the RPE reference plane) it would have been obtained by moving upward the cup-offset to a mean value of 201.6±53.5 microns.Discussion. Digital stereophotographs and StratusOCT showed a moderate agreement in assessing the VCDR. Measurements by the two techniques are not interchangeable in about one third of eyes, where the two techniques result in a difference higher than 0.1.Changing the position of the cup-offset on the basis of cup depth is likely to enhance the agreement.
Keywords: optic disc • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • imaging/image analysis: clinical