Premium
Description of the retinal vascular network by semi‐automated computer software (SIVA) in the MONTRACHET study
Author(s) -
Arnould L.,
Binquet C.,
Guenancia C.,
Alassane S.,
Kawasaki R.,
Daien V.,
Helmer C.,
Tzourio C.,
Bron A.M.,
CreuzotGarcher C.
Publication year - 2016
Publication title -
acta ophthalmologica
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.534
H-Index - 87
eISSN - 1755-3768
pISSN - 1755-375X
DOI - 10.1111/j.1755-3768.2016.0426
Subject(s) - retinal , tortuosity , vascular network , neurovascular bundle , ophthalmology , population , fractal dimension , medicine , anatomy , fractal , mathematics , engineering , mathematical analysis , geotechnical engineering , environmental health , porosity
Purpose The aim of this study was to identify retinal vascular network patterns using semi‐automated software in an elderly population over 75 years in a population‐based study, the MONTRACHET (Maculopathy Optic Nerve nuTRition neurovAsCular and HEarT diseases) study. Methods In 2009–2011, participants from the 3 Cities Study living in Dijon, were proposed to take part in the MONTRACHET study with a complete eye examination. During these exams, color retinal photographs were performed (centered on the optic disc). Semi‐automated software (Singapore Institute Vessel Assessment: SIVA) was used in order to describe the retinal vascular network of subjects with a photograph of sufficient quality. All subjects with epiretinal membranes and vascular pathology were excluded. Fifty‐four geometric features such as fractal dimension, diameter, tortuosity and branching angle were systematically collected. A principal component analysis (PCA) was used to identify independent patterns. Results Overall, 1,067 photographs were reviewed with SIVA software among the 1,153 participants of the MONTRACHET study. The mean age was 79.8 ± 3.8 years and 62.7% were female. After PCA processing, we extracted 3 vascular patterns summarizing 41% of the retinal vascular network information. The first pattern corresponded to a lower density in the vascular network as well as a higher variability in the vascular width (29% of the total variability). The second pattern was highly correlated to large vessels diameters and width. The third pattern matched to an increased tortuosity. Conclusions In this preliminary study, three vascular patterns were identified: decreased vascular density and heterogeneous network, increased vessels diameter and tortuosity. The next step of this study is to identify associations between these patterns and cardiovascular diseases in this population‐based study.