Purchase this article with an account.
R.F. Monteiro, R.C. Bernardes, E.S. Almeida, C.L. Lobo, J.G. Cunha-Vaz; Reability of the RTA for Detecting Changes in Retinal Thickness in Preclinical Retinopathy in Subjects With Type 2 Diabetes . Invest. Ophthalmol. Vis. Sci. 2003;44(13):3992.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: To assess the failure rate of the RTA (Retinal Thickness Analyzer, Talia, Israel) in evaluating the retinal thickness in patients with preclinical diabetic retinopathy (level 10 of Wisconsin ETDRS gradings). Methods: 26 eyes from 13 patients with diabetes type 2 and with preretinopathy stage were evaluated using the RTA software version 2.11 and analysis 6.35 over a 12 month period on 3 separate visits six-month apart. The patient's age ranged from 47 to 62 years old (mean: 56.2; SD: 5.4). One eye was excluded due the presence of cataract. First, the maps of thickness were inspected for the presence of abnormal shapes and their respective areas (fixating points, 1 to 9) were considered suspicious. Second, these suspicious areas were analyzed by inspecting the fundus image and their respective slit images. Third, the retinal thickness maps were compared against the reference map and the areas of detected increased thickness were matched with previously identified suspicious areas. Results: Of the total 675 areas considered (25 eyes x 3 visits x 9 fixating points), there were 101 suspicious areas (15%). Twelve of these areas presented poor fundus image quality while 51 presented poor slit image quality, representing, respectively, 1.8% and 7.6% of the total scanned areas. From the 101 suspicious areas, 78 matched positively with detected increased retinal thickness, which represents 11.6% of the possible false detections on detecting macular edema. Monitoring for both fundus and slit image quality, the error rate of possible false detections can be decreased to only 4.6%. Conclusions: Continuous monitoring image quality for on the RTA, both on fundus image and on slit images, can decrease the error rate in detecting macular edema from 11.6% to 4.6%.
This PDF is available to Subscribers Only