April 2011
Volume 52, Issue 14
Free
ARVO Annual Meeting Abstract  |   April 2011
A Web-based System for Vessel Tortuosity Estimation in Retinal Images
Author Affiliations & Notes
  • Alfredo Ruggeri
    Dept of Information Engineering, University of Padua, Padua, Italy
  • Enea Poletti
    Dept of Information Engineering, University of Padua, Padua, Italy
  • Diego Fiorin
    Dept of Information Engineering, University of Padua, Padua, Italy
  • Lara Tramontan
    Dept of Information Engineering, University of Padua, Padua, Italy
  • Tunde Peto
    Research & Development, Moorfields Eye Hospital, London, United Kingdom
  • Footnotes
    Commercial Relationships  Alfredo Ruggeri, None; Enea Poletti, None; Diego Fiorin, None; Lara Tramontan, None; Tunde Peto, None
  • Footnotes
    Support  NIHD BMRC for Ophthalmology
Investigative Ophthalmology & Visual Science April 2011, Vol.52, 2890. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Alfredo Ruggeri, Enea Poletti, Diego Fiorin, Lara Tramontan, Tunde Peto; A Web-based System for Vessel Tortuosity Estimation in Retinal Images. Invest. Ophthalmol. Vis. Sci. 2011;52(14):2890.

      Download citation file:


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

      ×
  • Supplements
Abstract
 
Purpose:
 

To develop a computerized web-based system to assess the vascular tortuosity of retinal images acquired from fundus camera. It will allow clinicians to recover the quantitative estimation of this parameter in an objective and user-friendly way.

 
Methods:
 

At first, a totally automated phase performs green channel extraction, image luminosity equalization and contrast enhancement, tracing of the vascular network, and vessel classification as either arteries or veins. In the second phase, the user evaluates these results and either confirms them or applies the necessary corrections via a user-friendly editing interface, which is capable of detecting the questionable situations and present them to the user for manual expert assistance. A third, totally automated phase computes the tortuosity score of each detected vessel, according to a previously published criterion, and finally combines these values, with a custom weight-averaged procedure, to derive a whole image tortuosity score. Accuracy of the complete system was assessed on 26 images manually classified over a range of four classes by Moorfields Eye Hospital retina experts, and on 48 images manually ordered by the same experts by increasing vessel tortuosity.

 
Results:
 

The association between the manual and automated classification was good, with a Pearson correlation coefficient of 0.92 for the 26-images set and a Spearman rank correlation of 0.81 for the 48-images set. The total time required to analyze one image was on average 5 min (range 1-12 minutes).

 
Conclusions:
 

A computerized system for the reliable quantification of vascular tortuosity in retinal images has been developed as a web-based system and will be soon available for general use.  

 
Keywords: imaging/image analysis: clinical • retina 
×
×

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.

×