June 2015
Volume 56, Issue 7
ARVO Annual Meeting Abstract  |   June 2015
High resolution imaging of retinal vascular network geometry
Author Affiliations & Notes
  • Jonathan Benesty
    CHNO des Quinze Vingts, Paris, France
    CIC 503 INSERM, Paris, France
  • Edouard Koch
    CIC 503 INSERM, Paris, France
  • David Rosenbaum
    Pitié-Sapetrière Hospital, Paris, France
  • Xavier Girerd
    Pitié-Sapetrière Hospital, Paris, France
  • José Sahel
    CHNO des Quinze Vingts, Paris, France
    CIC 503 INSERM, Paris, France
  • Florence Rossant
    Institut Supérieur d'Electronique de Paris, Paris, France
  • Isabelle Bloch
    Telecom ParisTech, Paris, France
  • Michel Paques
    CHNO des Quinze Vingts, Paris, France
    CIC 503 INSERM, Paris, France
  • Footnotes
    Commercial Relationships Jonathan Benesty, None; Edouard Koch, None; David Rosenbaum, None; Xavier Girerd, None; José Sahel, Imagine Eyes (S); Florence Rossant, None; Isabelle Bloch, None; Michel Paques, C (C)
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science June 2015, Vol.56, 5296. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jonathan Benesty, Edouard Koch, David Rosenbaum, Xavier Girerd, José Sahel, Florence Rossant, Isabelle Bloch, Michel Paques; High resolution imaging of retinal vascular network geometry. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):5296.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Purpose: An energetically optimal vascular network should theoretically respect some physical principles among which the cubic mathematical relationship between parent and daughter vessels at bifurcations (Murray’s law). Deviations from this optimal geometry have been reported in several vascular diseases. The availability of high resolution imaging of vessels offers the opportunity to reevaluate these results. Here, following our initial evaluation of adaptive optics (AO) retinal vascular imaging (Koch et coll. 2014), we mathematically analyzed arterial and venous bifurcations to calculate more precisely the junction exponent and to evaluate the effect of cardiovascular risk factors.

Methods: Flood-illumination AO imaging was done in 93 arterial and 25 venous bifurcations of 21 controls and 65 patients (mean age 45.7 years ±15.2; M/F 44/49) with various cardiovascular risk factors (hypercholesterolemia, hypertension, diabetes, smoking). A semi-automated procedure was used to measure vascular diameters and angles.

Results: In control arteries, the mean value of the arterial junction exponent (theoretically equal to 3) was 2.80 (±0.44). The linear regression for measured vs predicted parent arterial diameter was 0.993 with a coefficient of determination R2=0.942. Mean arterial bifurcation angle measured was 85.05° (±11.7°), which was significantly different from the mean optimal angle 71.7°(±13.9°) (p<0.01). In the cardiovascular risk group, the arterial junction exponent was significantly higher than controls (3.24±1.4 versus 2.8±0.44; p=0.02) but not the venous junction exponent (2.47 ±0.45 versus 2.42 ±0.46; p=0.8). The junction exponent in the diabetic group (n=9) was 3.9 (±2.5), no difference was found with the control group (p=0.25) but it might be a consequence of the small number of diabetic patients.

Conclusions: In arterial bifurcations of normal eyes, the junction exponent is not significantly different from 3, further confirming Murray's law as a general principle for small arteries but not veins. Patients with cardiovascular risk factors have a higher arterial junction exponent, suggesting an alteration of flow distribution. AO imaging of arterial bifurcations may therefore provide a novel biomarker of the general vascular state. Further studies are needed to understand the anatomical, functional and clinical significance of this finding.


This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.