A new method for visualizing drusen and their progression in adaptive optics ophthalmoscopy

Abstract

Purpose : Drusen, a hallmark of age-related maculopathy, are dynamic, mainly sub-RPE deposits whose progressive increase in size are predictive of evolution to late stages of age-related macular degeneration (AMD). Here we propose a new method for visualizing drusen and their progression using adaptive optics ophthalmoscopy (AOO). This approach takes advantage of the fact that the contrast of drusen varies with gaze position in flood-illumination adaptive optics (FIAO) ophthalmoscopy. We capitalize on this to show that by mathematically combining images from different gaze positions, drusen microstructure can be revealed with high resolution and contrast.

Methods : AMD patients (n=41) and controls (n=15) were recruited from the Quinze-Vingts National Ophthalmology Hospital and University of Pittsburgh Department of Ophthalmology. A FIAO fundus camera (rtx1, ImagineEyes, France) was used to acquire a 4°×4° fovea-centered image and then its internal fixation target was moved to obtain 4–8 additional, overlapping images, with gaze displaced by +/- 2° vertically and horizontally (or both). Custom software registered images and calculated the standard deviation (SD) of each pixel across areas of overlap. Longitudinal imaging was done in 27 patients every 3–12 months for up to 3 years.

Results : No gaze-varying structures were revealed in controls. In patients, drusen as small as 30 µm were detected with high contrast. The smallest appeared as uniformly bright (denoting a high SD). Larger ones had bright margins with a dark center and the largest sometimes had additional bright structures within the dark center. Individual drusen were clearly delineated within densely packed drusen. In two patients with normal vision and no evidence of drusen on OCT, diffuse subretinal opacities were detected. Time-lapse images revealed several patterns of drusen progression, including expansion of drusen and fusion of drusen to form larger aggregates.

Conclusions : Gaze dependent variability in FIAO images of drusen permits contrast enhancement to visualize drusen microstructure and progression with high contrast and resolution. This new technique appears to be a robust, sensitive method to detect, map, measure, and monitor age-related drusen progression. This type of visualization may be particularly useful to document the earliest stages of aging of the retina-RPE complex, and to document the dynamics of drusen progression in AMD.

This is a 2020 ARVO Annual Meeting abstract.