Evaluation of optic disc cupping using high-resolution ocular ultrasound.

Abstract

The new Biovision B-Scan Probe has greater resolution than its predecessors because of its posteriorly placed focus (25 mm inside the eye) and the consequent reduction in size of the sound field. We investigated the accuracy, reproducibility and test-retest variability of two-dimensional optic cup measurements using this new-generation B-scanner. One randomly selected eye of 20 patients underwent five repeated measurements of vertical (VOCD) and horizontal optic cup diameters (HOCD), and optic cup depth, using confocal scanning laser tomography (Heidelberg Retinal Tomograph) and high-resolution ultrasound (Biovision, Chiron Vision, UK) on two separate occasions. There was one operator per instrument who was masked to all clinical data and previous measurements. Accuracy of ultrasound readings was quantified by comparing the results with those of Heidelberg Retinal Tomography (HRT). The 95% confidence interval for the bias of echographic vertical and horizontal cup diameter measurements was -95 +/- 48 microns and -19 +/- 72 microns respectively, and -87 +/- 328 microns for cup depth readings. The indices of reproducibility (mean coefficient of repeatability +/- SD) were: VOCD: 177 +/- 105 microns [B-scan] 209 +/- 100 microns [HRT]; HOCD: 179 +/- 61 microns [B-scan], 205 +/- 101 microns [HRT]; cup depth: 206 +/- 63 microns [B-scan], 204 +/- 124 microns [HRT]. Ninety-five per cent limits of agreements between initial and retest values for ultrasound were 18 +/- 136 microns [VOCD], 2 +/- 144 microns [HOCD] and 4 +/- 156 microns [cup depth], compared with 1 +/- 104 microns [VOCD], 20 +/- 102 microns [HOCD] and 3 +/- 168 microns [cup depth] for scanning laser tomography. The results demonstrate that measurements of optic cup diameter and depth using ultrasound correlate strongly with corresponding HRT readings. These echographic measurements are reproducible, and not subject to clinically meaningful test-retest variability. This technique of measuring two-dimensional cup parameters does not require expensive specialist equipment and has many potential clinical applications which are discussed.