Related Experiment Video
Updated: Oct 1, 2026

Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
Retinal Vascularity Index in Mild Cognitive Impairment
Rachel D'Cunha1, John Hsu1, Jenny Melcher2
1iMIND Study Group, Department of Ophthalmology, Duke University School of Medicine, Durham, NC, USA.
Purpose:
The purpose of this study was to determine whether retinal vascularity index (RVI) differs between individuals with mild cognitive impairment (MCI) and normal cognition, and to assess RVI as a novel ultra-widefield (UWF) implementation of retinal arteriovenous assessment.
Methods:
In a cross-sectional case-control study, UWF fundus color images from patients with MCI and cognitively healthy controls were analyzed using a semi-automated platform via the Comprehensive Ocular Imaging Network (COIN; ocularimaging.net). RVI was calculated as the ratio of arterial-to-venous cross-sectional area within a 47-degree region of interest (ROI) centered on the optic disc, with quadrant-based segmentation producing regional RVI values for 4 quadrants within the 47-degree ROI. Generalized estimating equations (GEEs) accounted for inter-eye correlation.
Results:
Fifty-five eyes of 38 patients with MCI and 89 eyes of 55 cognitively healthy patients were included. Mean RVI across the 47-degree ROI was reduced in MCI versus control eyes (0.524 vs. 0.730, P < 0.001) after adjusting for treated hypertension. Reduced RVI was also observed in each of the four quadrants in MCI compared with controls (P < 0.001). RVI showed good discriminative ability (area under the curve [AUC] = 0.86, sensitivity = 80%, and specificity = 79%) with good inter-grader reliability (intraclass correlation coefficient [ICC] = 0.840, 95% confidence interval [CI] = 0.686-0.943).
Conclusions:
Mean RVI across the 47-degree ROI was significantly lower in individuals with MCI compared with cognitively healthy controls.
Translational Relevance:
RVI, derived from UWF imaging, may reflect retinal microvascular cross-sectional area differences associated with MCI. As an exploratory metric, further validation is needed before RVI can be used as a noninvasive tool to complement diagnostic approaches for MCI.
