Purchase this article with an account.
David Perlmutter, Christopher J Brady, Omesh Gup, Krista Dedrick, Jay H Federman; Retinal Vascular Measurements in Pre-Hypertension and Pre-Diabetes: Validation of Instrument. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5356.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To obtain and validate baseline central retinal arterial equivalent (CRAE), central retinal vein equivalent (CRVE), and retinal arterio-venous ratio (AVR) measurements in a clinical patient population with pre-diabetes mellitus and pre-hypertension using the AVRnet software.
The Canon DGi Digital Imaging System was used to take non-mydriatic fundus photos centered on the optic disc in a Family Medicine research office. Sixty-four images were then transmitted via a teleophthalmology interface, and analyzed remotely with AVRnet software to obtain CRAE, CRVE, and AVR measurements. Twenty sequential images were regraded by the orginal grader to assess intragrader reliability using intraclass correlations (ICC). The same 20 images were also graded by 2 independent graders to assess intragrader reliability.
Mean CRAE was 146.2 (standard error (SE)=1.697). Mean CRVE was 226.9 (SE=2.532). Mean AVR was 0.648 (SE=0.0087). Individual ICC for the intragrader assessment of CRAE was 0.695, of CRVE was 0.943, and of AVR was 0.787. Individual ICC for the intergrader assessment of CRAE was 0.77, of CRVE was 0.904, and of AVR was 0.765.
AVRnet is a reliable analytic tool for the measurement of the CRAE, CRVE, and AVR in non-mydriatic fundus photos. Values obtained showed acceptable intragrader and intergrader reliability, and were similar to published values in other populations. Correlations with other vasculopathic risk factors could provide prognostic information to primary care providers.
This PDF is available to Subscribers Only