Reliability of computer-assisted retinal vessel measurementin a population

Clin Exp Ophthalmol. 2002 Jun;30(3):179-82. doi: 10.1046/j.1442-9071.2002.00520.x.


The purpose of the study was to assess the intergrader and intragrader reliability of computer-assisted retinal vessel dia-meter measurement in a defined, community-based population. Retinal photographs from participants in the Blue Mountains Eye Study were digitized using standard techniques. A grader identified all retinal vessels located 0.5-1.0 disc diameter from the optic disc margin,and a computer program measured the width of these vessels. Intergrader and intragrader reliability was assessed on a random sub-sample of 184 and 97 images, respectively, using quadratic weighted kappa(kappa) and correlation analysis (R2). Intergrader reliability was high for summary indices of retinal arteriolar (kappa = 0.85, R2 = 0.88)and venular (kappa = 0.90, R2 = 0.90)diameters, and their ratio, the arteriole-to-venule ratio (kappa = 0.75, R2 = 0.79).Intragrader reliability was also high, with kappa values ranging from 0.80 to 0.93 and from 0.80 to 0.92 for graders 1 and 2, respectively. It is concluded that the retinal vessel diameters could be reliably measured using computer-assisted software and may be used for population-based research.

Publication types

  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Female
  • Humans
  • Image Processing, Computer-Assisted / methods
  • Image Processing, Computer-Assisted / standards*
  • Male
  • Middle Aged
  • Observer Variation
  • Reproducibility of Results
  • Retinal Vessels / anatomy & histology*