Purchase this article with an account.
Kyungmoo Lee, Ian C Han, Douglas B. Critser, Andreas Wahle, Milan Sonka, Michael David Abramoff; Layer-Specific Vascular Analysis of OCT Angiography in Diabetic Macular Edema. Invest. Ophthalmol. Vis. Sci. 2017;58(8):644.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
OCT angiography (OCTA) is a new non-invasive imaging technique used for diagnosing and monitoring retinal vascular diseases. Quantification of layer specific capillary metrics is of high value to a better understanding and management of diabetic macular edema (DME). However, layer segmentation algorithms typically fail in DME with substantial retinal thickening, intercellular fluid and cysts. The aim of this study is to evaluate the Iowa Reference Algorithms and layer specific OCTA.
Eleven OCT and OCTA volume pairs (350 × 1024 × 350 voxels, 6.02 × 2.00 × 6.02 mm3) were obtained from 11 eyes (6 OD, 5 OS) of 11 patients (54.5 ± 11.1 years) with DME using a Cirrus 5000 HD-OCT Angioplx (Carl Zeiss Meditec Inc., Dublin, CA). Ten retinal layers were automatically segmented from the OCT volume using the Iowa Reference Algorithms (http://www.iibi.uiowa.edu/content/shared-software-download, Figs. 1ab) and defined the retinal layers in the OCTA volume (Figs. 1cd). For each layer, an OCTA density map was created by averaging voxel intensities in the depth direction to identify vasculature (Figs. 1eg). A capillary density map was generated by segmenting the smallest vessel and determining density (Figs. 1fh).
Fig. 1 shows a typical example for our layer-specific vascular analysis of OCT angiography. The segmented inner retinal layers are located above the fluid-filled regions (Fig. 1b). In 10 / 11, these maps were obtained successfully.
Quantification and analysis of vasculature for layer specific OCTA volumes are possible using the proposed method, in most cases, and have the potential to increase our understanding and management of diabetic macular edema, an important cause of blindness.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
Figure 1. (a) B-scan of the OCT volume with DME. (b) Panel (a) layer segmentation. (c) B-scan of the OCTA volume. (d) Panel (c) with layer segmentation. (e, f) OCTA density map and capillary density map of retinal nerve fiber layer plus ganglion cell layer (RNFL+GCL). (g, h) OCTA density map and capillary density map of inner plexiform layer plus inner nuclear layer (IPL+INL).
This PDF is available to Subscribers Only