Purchase this article with an account.
S.J. Morgan, J. Lusty, S. Hatcher, L. Beck, D. Owens; A Comparison of Two Digital Image Capture Cameras with Slit-lamp Bio-microscopy in the Grading of Diabetic Retinopathy . Invest. Ophthalmol. Vis. Sci. 2003;44(13):4030.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: To compare in a clinic environment the effect of digital image capture camera resolution upon the grading of diabetic retinopathy. Methods: 50 Eyes from 25 diabetic patients were photographed using two digital camera systems mounted on a Canon CR6-45NMf Retinal Camera. The cameras used were the Sony DXC-950P Video Camera (752 x 582 pixels, 1.3Mb file size) and the Canon D30 Digital Camera (2160 x 1440 pixels, 9Mb file size). Mydriatic fundus photography comprised of two 45 degree fields (one macular centred and one nasal to disc) was obtained for each eye and the images stored in their uncompressed format. These images were then graded using the Bro Taf Diabetic Retinopathy Screening Service criteria. Each patient also underwent dilated indirect fundoscopy for independent grading by a single Ophthalmologist within one week of photography. The grading results from the two cameras and slit-lamp bio-microscopy were then compared. Results: Slit-lamp bio-microscopy grading agreed with the Canon D30 digital camera image grading in 36% of eyes compared to 46% agreement with the Sony 950 Video Camera. The ophthalmologist grading of the retinopathy was higher (more clinically advanced) in 46% and 40% of eyes compared to the Canon D30 and Sony 950 images respectively. The slit-lamp bio-microscopy graded the retinopathy as less severe in 18% of eyes compared to the Canon D30 and 14% of eyes compared to the Sony 950. Further statistical analysis will be presented. Conclusions: The availability of higher resolution images of diabetic retinopathy did not increase the agreement between the gradings from the digital image capture cameras studied, and slit-lamp bio-microscopy. This has implications for the imaging standards to be used in future community based diabetic retinopathy screening programs.
This PDF is available to Subscribers Only