personal scheduler Hamburg 2013 - Main Site Press Release 2013 Call for Abstracts General Information Programme Overview Letter from the President Keynote Lectures Main Sessions NEW - Free Paper Session Instructional Courses World Retina Day Retinal Detachment Course Uveitis Course Other Symposia Surgical Skills Training Courses
hamburg banner

New software to assess retinal non-perfusion on Optomap? wide-field fundus fluorescein angiography in diabetic macular oedema (DMO)

Session Details

Session Title: Vascular Diseases I

Session Date/Time: Friday 27/09/2013 | 08:00-10:00

Paper Time: 09:28

Venue: Hall C (Level 1)

First Author: P.Stanga UK

Co Author(s):    S. Caputo   H. Jaberansari   Y. D?Souza   D. Henson     

Abstract Details


To develop new software that takes into account image distortion to assess the location and quantify the area of retinal non-perfusion (RNP) and its relationship with Central Macular Thickness (CMT) in patients with Diabetic Macular Oedema (DMO) on wide-field fundus fluorescein angiography (WF-FFA).


Vitreomacular clinic


Retrospective review of 76 Optomap® WF-FFA images of patients with DMO. A grid was superimposed over the WF-FFA image (each cell being one disc area of 1.77mm2). The image was divided into 3 zones: 1.Posterior Pole (PP): within an ellipse centered on the fovea, passing through a point one disc diameter from the nasal edge of the optic disc (OD) and including the vascular arcades; 2 Mid-peripheral (MP): outside the PP but within a circle centered on the OD and passing along the posterior edge of the vortex vein ampullae on the green-free image; 3.Peripheral (P): anterior to the MP zone. Two independent graders classified each cell within zones 2 & 3 as: 1(>50% perfused), 2(<50% perfused), 3(transition area between 1 and 2), 4(poor image quality) or 5(blockage by eyelashes/eyelids). In case of disagreement of >10 cells scored as 2 or 3 in zone 2 the image was rescored by the two graders together. The coefficient of agreement was used to determine the consistency between the two graders initial scores. Correlation coefficients were calculated to assess the relationship between CMT and area of RNP.


Coefficient of agreement for the MP was 21 cells. Thirteen images were regraded due to differences >10cells between the 2 graders. Mean CMT was 290µ(range:199-606µ). Average of RNP areas in MP was 4.17(range:0-25). Correlation coefficient with CMT was 0.349, p=0.019. Average of RNP areas in P was 2.01(range:0-29). There was no correlation with CMT and peripheral RNP areas.


This new software package allows for an easier and faster quantification of the area and location of retinal non-perfusion in the mid and periphery on Optomap® WF-FFA images while, for the first time, taking into account the distortion that occurs in WF-FFA images. Retinal non-perfusion associated with DMO is mainly located in the mid-periphery. We found a weak correlation between CMT and mid-peripheral retinal non-perfusion. Further studies using this new software are required to ascertain the area and role of retinal non-perfusion in DMO and other ischemic retinal diseases.

Back to Freepaper Session
EURETINA, Temple House, Temple Road, Blackrock, Co Dublin. | Phone: 00353 1 2100092 | Fax: 00353 1 2091112 | Email:

Privacy policyHotel Terms and Conditions Cancellation policy