Purchase this article with an account.
Zuzana Sipkova, Anthia Papazoglou, David Sculfor, Consuela Moorman, Larry Benjamin, Peter Bankhead, Graham McGeown, Timothy Curtis, Victor Chong; The Use Of A Novel Software In Predicting Progression Of Diabetic Retinopathy Through Measurement Of Retinal Vascular Changes. Invest. Ophthalmol. Vis. Sci. 2012;53(14):2844.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate a new software in measuring retinal vascular changes in patients with diabetes as a risk factor for progression.
The database of a hospital-based diabetic retinopathy screening was assessed retrospectively. 408 patients were identified to have over 6 years of follow up between 2004 and 2011. Patients with retinopathy on baseline were excluded from the study. Non-Progression group (NPG) is defined as no retinopathy on baseline or at last follow up, whilst Progression group (PG) had no retinopathy on baseline but retinopathy present at last follow up. We have used only the right eye of each patient for analysis. In this preliminary study, we have selected a random sample of 35 images from each group. The central retinal arteriolar (CRAE) and venular equivalence (CRVE) on baseline images were measured using a new software. Using this software, the CRAE and CRVE can be calculated in patients with only the temporal part of the optic disc included in the image. Two observers (ZS and AP) have carried out the measurement independently. Student t-test was used for statistical analysis.
The inter-observer differences were not statistically significant. The mean measurement was used for analysis. Out of the 70 patients, 14 (7 from each group) were excluded due to poor image quality. The mean CRAE unit in the NPG and PG were 14.27 (SD 1.67) and 14.88 (SD 1.36), respectively, and the mean CRVE unit in the NPG and PG were 17.85 (SD 1.67) and 18.90 (SD 1.98). The progression group has significantly wider CRVE (p=0.018) but not CRAE (p=0.068) than the non-progression group. The A-V ratio did not show a significant difference (p=0.31).
This new software appears to be able to measure retinal vascular changes reliably. We were able to confirm even with this small sample size that dilation of retinal vein is a risk factor of diabetic retinopathy progression. The full dataset analysis is currently under evaluation.
This PDF is available to Subscribers Only