Purchase this article with an account.
Zhihong Hu, Amir H Hariri, Srinivas R Sadda; Directional Progression Analysis of Geographic Atrophy in Fundus Autofluorescene Images. Invest. Ophthalmol. Vis. Sci. 201657(12):.
Download citation file:
© 2017 Association for Research in Vision and Ophthalmology.
Geographic atrophy (GA) is the late-stage of age-related macular degeneration (AMD). The purpose of this study is to test the hypothesis that the expansion speed of GA is associated with its direction of growth relative to the foveal center.
Thirty one eyes from 31 subjects (one eye per subject) with evidence of GA were included in this study. Each eye underwent fundus autofluorescence (FAF) imaging (Spectralis HRA+OCT) at baseline and 18 months later. The FAF image resolution varied from 496 x 596 pixels to 1536 x 1536 pixels with a physical size of 9 mm x 9 mm. All the images were resized to a consistent dimension. Manual delineation of GA regions was performed on the 62 FAF images from the two visits of the 31 eyes. A feature-based image registration algorithm was applied to align the FAF images of the two visits. The corresponding manually-delineated GA masks were also aligned using the same transformation. The GA masks were converted to binary images and thresholded using Ostu's global thresholding. GA and hole boundaries were detected based on 8-connectivity. Directional GA growth rate was defined as the GA growth along the normal direction of each GA boundary point from the baseline visit to the month 18 visit. GA lesions which merged together were excluded from the GA growth analysis.
Fig. 1 illustrates the GA boundaries and Fig. 2 shows the directional GA progression result. For all the 31 eyes, the GA lesions grow towards the foveal center with a mean of 0.144 ± 0.188 mm/year which is significantly less than the growth rate away from the foveal center with a mean of 0.211 ± 0.248 mm/year (p<0.01, unpaired T-test).
The directional progression of GA was faster when growing away from the foveal center than growing towards the foveal center. These findings may have implications for assessing GA growth in the context of therapeutic trials.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
Fig.1. FAF images of baseline (upper left) and month 18 (upper right) and their corresponding GA gradings (bottom). Note: green lines demarcate the outer GA boundaries and red lines denotes the hole boundaries due to foveal sparing.
Fig.2. GA progression result. Note: red dots represent the GA directional growth rate at each individual boundary point from the baseline visit. Blue and cyan colors indicate the mean growth rate and standard deviation for all the boundary points of all the 31 eyes.
This PDF is available to Subscribers Only