April 2011
Volume 52, Issue 14
Free
ARVO Annual Meeting Abstract  |   April 2011
Automated Image Alignment to Follow Progression of Geographic Atrophy in Age-Related Macular Degeneration
Author Affiliations & Notes
  • David J. Ramsey
    Ophthalmology,
    Johns Hopkins Wilmer Eye Inst, Baltimore, Maryland
  • Janet S. Sunness
    Hoover Rehab Low Vision Svc, Greater Baltimore Med Ctr, Baltimore, Maryland
  • James T. Handa
    Johns Hopkins Wilmer Eye Inst, Baltimore, Maryland
  • Footnotes
    Commercial Relationships  David J. Ramsey, None; Janet S. Sunness, None; James T. Handa, None
  • Footnotes
    Support  Unrestricted award from Research to Prevent Blindness (Wilmer), NIH EY08552 (JS), and the Robert Bond Welch Professorship (JH).
Investigative Ophthalmology & Visual Science April 2011, Vol.52, 4470. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      David J. Ramsey, Janet S. Sunness, James T. Handa; Automated Image Alignment to Follow Progression of Geographic Atrophy in Age-Related Macular Degeneration. Invest. Ophthalmol. Vis. Sci. 2011;52(14):4470.

      Download citation file:


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

      ×
  • Supplements
Abstract
 
Purpose:
 

Progression of geographic atrophy (GA) associated with age-related macular degeneration is often monitored using fundus photographs. Autofluorescence imaging is often inadequate for delineation of parafoveal atrophy because of absorption of blue light by xanthophyll. Many variables affect the utility of photographic records as a tool for measuring the growth of GA, ranging from differences in image alignment and focus, to patient-specific factors such as the ability of a patient with declining visual function to reliably fixate. The present study evaluates the utility of a commercially available image alignment software package, i2k Align Retina, to correct for differences in image alignment, magnification, and other optical aberrations in order to improve the accuracy of clinicians to follow the progression of GA.

 
Methods:
 

A longitudinal series of 30 and 60 degree color fundus photos from patients with GA and foveal function at baseline from the Wilmer Geographic Atrophy Study were included in the analysis. The areas of GA were identified, measured, and the rate of atrophy enlargement assessed both with and without image post-processing until foveal function was lost, as judged by a decline in visual acuity to below 20/60.

 
Results:
 

Of the 65 patients with foveal function at baseline (median visual acuity 20/34), 64 patient data sets could be aligned by i2k Align Retina software (98%). Foveal function persisted for an average of 2.5±1.6 yrs, providing a total of 212 images for analysis. The average distortion correction by image post-processing was 1.0±5.4%. The absolute value of the distortion was 4.1±3.7% with a maximum correction of 13.6% of image area. A majority of the distortion arose from the alignment of 30 with 60 degree fundus photos (absolute value of image distortion 7.7±3.0%). No difference in the rate of atrophy enlargement was detected between the pre- and post-processed images (2.66 mm2/yr versus 2.73 mm2/yr, N.S.).

 
Conclusions:
 

The digital revolution in fundus imaging provides an opportunity to apply image post-processing to enhance quantitative and qualitative assessment of retinal disease. Image alignment allows for improved visual analysis of GA progression by precisely aligning longitudinally collected fundus photos. This automated software platform also enhances the integration of older photographic records with those obtained using newer digital imaging systems. Using this technology, fundus features that predict the progression of GA are currently under investigation.

 
Keywords: age-related macular degeneration • imaging/image analysis: clinical • image processing 
×
×

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.

×