Purchase this article with an account.
Alauddin Bhuiyan, Uyen Nguen, Ryo Kawasaki, Ecosse Luc Lamoureux, Paul Mitchell, Kotagiri Ramamohanarao, Tien Y Wong, Yogesan Kanagasingam; Inter- and Intra-observer agreement of Newly Developed Semi-automated Software for Retinal Artery-Vein Nicking Quantification and Severity Grading. Invest. Ophthalmol. Vis. Sci. 2014;55(13):234.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess the reliability of novel computer based semi-automated software for quantifying the severity of retinal artery-vein nicking (AVN), a hypertensive retinopathy sign, from colour retinal photographs.
47 retinal photographs were randomly selected from the Singapore Malay Eye Study (SiMES), a population-based cross-sectional study of persons aged 40-80 years. A novel computer assisted software was developed to quantify retinal AVN by selecting vessel crossover region semi-automatically. The software then calculates the AVN values automatically. The accuracy was tested against the manually measured AVN severity level by expert grader through the Spearman rank correlation. Inter-grader and intra-session reliability were assessed by measurements from two individual graders and by grading two images of the same eye taken at different times, using Intra-class Correlation Coefficients (ICC).
Accuracy measurements indicated a Spearman rank correlation (ρ) of 0.71 (p<0.01) between software measured AVN values and expert graders’ ranking on AVN severity. Between-grader and intra-session ICC value for AVN was 1, which indicates high repeatability of the proposed software for AVN severity measurement.
A novel semi-automatic measurement system for retinal artery vein nicking quantification is demonstrated for the first time. A good agreement with standard grading and excellent reproducibility, indicating that quantitative assessment of retinal AVN is possible. This can lead to achieve higher accuracy on CVD prediction model and early diagnosis of ocular diseases.
This PDF is available to Subscribers Only