Drusen detection by confocal aperture-modulated infrared scanning laser ophthalmoscopy

摘要

Aim To evaluate the efficiency of drusen detection by scanning laser ophthalmoscopy (SLO) using various infrared confocal apertures and differential contrast (DC) strategies. Methods 11 eyes with non-neovascular age-related macular degeneration (AMD) underwent infrared imaging with a Nidek F-10 confocal SLO using multiple confocal apertures: central, ring, aperture on the right side (AR) and left side (AL), with and without use of the DC. A conventional colour fundus photograph was also obtained. Images were exported into a certified grading tool and all visible drusen were manually outlined by two graders. For each image type, the number of drusen and total drusen area were calculated, and the measurements obtained by the two graders were averaged. Intergrader reliability was evaluated, and paired t tests compared measurements between the various aperture/DC modes and the colour image. Results Agreement between graders was high (r=0.93-0.98). Drusen number values obtained with the AR (121.0, p=0.01) mode were higher than for the colour photographs (69.1). Area measurements were also significantly higher in the AR (1.93 mm2; p=0.04) and AL modes (1.41 mm2; p=0.03) when compared with the colour photographs (1.24 mm2). The addition of the DC did not seem to improve drusen detection compared with the unmodified infrared images. Conclusions In this pilot study, drusen number and area grades were significantly higher using the AR and AL in which the laterally scattered light is captured (retromode). Use of the lateral confocal aperture may highlight subclinical drusen and aid in monitoring disease progression and response to emerging nonneovascular AMD therapies.