June 2013
Volume 54, Issue 15
Free
ARVO Annual Meeting Abstract  |   June 2013
Novel method for registration and statistical analysis of volumetric retinal OCT data using surface currents
Author Affiliations & Notes
  • Evgeniy Lebed
    Engineering Science, Simon Fraser University, Burnaby, BC, Canada
  • Sieun Lee
    Engineering Science, Simon Fraser University, Burnaby, BC, Canada
  • Sherry Han
    Engineering Science, Simon Fraser University, Burnaby, BC, Canada
  • Paul Mackenzie
    Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, BC, Canada
  • Alain Trouve
    Centre de Mathématiques et Leurs Applications, Ecole Normale Supérieure, Cachan, France
  • Marinko Sarunic
    Engineering Science, Simon Fraser University, Burnaby, BC, Canada
  • Mirza Faisal Beg
    Engineering Science, Simon Fraser University, Burnaby, BC, Canada
  • Footnotes
    Commercial Relationships Evgeniy Lebed, None; Sieun Lee, None; Sherry Han, None; Paul Mackenzie, None; Alain Trouve, None; Marinko Sarunic, None; Mirza Faisal Beg, None
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science June 2013, Vol.54, 5534. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Evgeniy Lebed, Sieun Lee, Sherry Han, Paul Mackenzie, Alain Trouve, Marinko Sarunic, Mirza Faisal Beg; Novel method for registration and statistical analysis of volumetric retinal OCT data using surface currents. Invest. Ophthalmol. Vis. Sci. 2013;54(15):5534.

      Download citation file:


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

      ×
  • Supplements
Abstract
 
Purpose
 

Volumetric imaging with OCT is being used to study the possible role of the ONH morphometry and changes in glaucoma. Registration of volumetric data is crucial for longitudinal or cross-sectional analysis of degeneration and morphological changes that may precede vision loss. In this report, we present a novel method to simultaneously register multiple surfaces from the ONH and perform corresponding vertex-wise biomarker (thickness) comparisons.

 
Methods
 

Six control eyes and five glaucomatous eyes were imaged using a custom-built 1060nm swept source OCT system. Using graph-cuts based segmentation, we extracted 4 surfaces: ILM, NFL, BM and choroid (CH), and computed the NFL and CH thickness for each subject. All of the surfaces were first registered to a common template (Fig. 1) by using the method of “surface currents.” The transformed template surfaces were subsequently further mapped to the target surfaces using “spherical demons”-based registration. This procedure gives a point-to-point homology between corresponding surfaces across different eyes, allowing for direct vertex-wise comparisons of surface-based biomarkers (e.g. choroidal and NFL thickness) between the control and glaucomatous group

 
Results
 

After performing a correction for multiple comparisons, we were able to identify regions, Fig. 2, on the template surface that contained statistically significant differences in the measurements of NFL thickness and CH thickness in the glaucomatous group relative to the control group.

 
Conclusions
 

We have presented a new method for simultaneously registering multiple surfaces extracted from volumetric OCT images to generate point-wise homology. The homology produced by this method allowed for a direct comparison of surface-based biomarkers at every vertex taken from two groups. The same method can be used to generate vertex-wise correspondence across longitudinal OCT morphometric data, allowing for analysis of point-to-point changes over time. This method represents a new tool for studying possible morphometric changes to the ONH in diseases such as early glaucoma.

     
Keywords: 549 image processing • 551 imaging/image analysis: non-clinical • 550 imaging/image analysis: clinical  
×
×

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.

×