June 2015
Volume 56, Issue 7
Free
ARVO Annual Meeting Abstract  |   June 2015
Automated Quantitative Analysis of Leakage and Ischemia for Ultra-widefield Angiography in Retinal Vascular Disease
Author Affiliations & Notes
  • Carley Tanchon
    Cole Eye Institute, Cleveland Clinic, Chagrin Falls, OH
  • Sunil K Srivastava
    Cole Eye Institute, Cleveland Clinic, Chagrin Falls, OH
  • Justis P Ehlers
    Cole Eye Institute, Cleveland Clinic, Chagrin Falls, OH
  • Footnotes
    Commercial Relationships Carley Tanchon, None; Sunil Srivastava, Allergan (R), Bausch and Lomb (C), Bausch and Lomb (R), Bioptigen (P), Synergetics (P); Justis Ehlers, Alcon (C), Bioptigen (C), Bioptigen (F), Bioptigen (P), Genentech (F), Leica (C), Leica (F), Synergetics (P), Thrombogenics (C), Thrombogenics (F), Zeiss (C), Zeiss (F)
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science June 2015, Vol.56, 3067. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Carley Tanchon, Sunil K Srivastava, Justis P Ehlers; Automated Quantitative Analysis of Leakage and Ischemia for Ultra-widefield Angiography in Retinal Vascular Disease. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):3067.

      Download citation file:


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

      ×
  • Supplements
Abstract
 
Purpose
 

To develop and establish an automated method for objective quantitative analysis of leakage and ischemia parameters in ultra-widefield angiography.

 
Methods
 

An IRB-approved retrospective review of patients with retinal vascular disease who had undergone ultra-widefield angiography by the Optos 200Tx System was performed. Two independent reviewers examined the ultra-widefield fluorescein angiograms (UWFA) for image quality and limited artifacts. Two images were selected for each eye, one early and one late. A computerized automated algorithm was developed to identify areas of leakage and ischemia with co-registration of the two image timepoints. Manual segmentation was also performed with Image J software for validation of the computerized algorithm. Quantification of ischemia was based on proportion of areas identified within the overall retinal area, “ischemic index.” Similarly, leakage was also evaluated for proportion of activity to develop a similar metric or “leakage index.”

 
Results
 

Thirty-two eyes of 32 patients with diabetic retinopathy or retinal venous occlusions (RVO) were included in this study. All images were successfully analyzed with the automated algorithm. All eyes (100%) displayed macular and/or peripheral leakage, while only 18 of 32 eyes presented with ischemia. Of the 32 eyes, the leakage index was a mean of 4.4% (range: 0.08-13%). Of the 18 eyes with ischemia, the mean ischemic index was 13.7% (range: 2.9-37.7%), most of which was peripheral.

 
Conclusions
 

UWFA in diabetic retinopathy and RVO can be analyzed with a computer-based segmentation algorithm to delineate areas of ischemia and leakage. Significant pathology is noted in the retinal periphery that may overlooked with standard angiography. Improving our understanding of the importance and relative location of these features (e.g., far/mid-periphery, macula) may be critical in understanding disease burden, treatment response, and prognosis. Further research is needed to enhance higher order identification of various types of leakage (e.g., perivascular, generalized, microaneurysms).  

 
Figure 1: Automated segmentation samples of diabetic retinopathy revealing leakage (green) and ischemia (blue).
 
Figure 1: Automated segmentation samples of diabetic retinopathy revealing leakage (green) and ischemia (blue).
 
 
Figure 2: Automated segmentation samples of RVO revealing leakage (green) and ischemia (blue).
 
Figure 2: Automated segmentation samples of RVO revealing leakage (green) and ischemia (blue).

 
×
×

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.

×