June 2015
Volume 56, Issue 7
Free
ARVO Annual Meeting Abstract  |   June 2015
Improved 3D retinal shape models with high-field MRI
Author Affiliations & Notes
  • Jan-Willem M Beenakker
    Department of Ophthalmology, Leiden University Medical Center, Leiden, Netherlands
    Department of Radiology, C.J. Gorter Center for High-field MRI, Leiden University Medical Center, Leiden, Netherlands
  • Denis P Shamonin
    Department of Radiology, Division of Image Processing, Leiden University Medical Center, Leiden, Netherlands
  • Andrew G Webb
    Department of Radiology, C.J. Gorter Center for High-field MRI, Leiden University Medical Center, Leiden, Netherlands
  • Berend C Stoel
    Department of Radiology, Division of Image Processing, Leiden University Medical Center, Leiden, Netherlands
  • Gregorius P M Luyten
    Department of Ophthalmology, Leiden University Medical Center, Leiden, Netherlands
  • Footnotes
    Commercial Relationships Jan-Willem Beenakker, None; Denis Shamonin, None; Andrew Webb, None; Berend Stoel, None; Gregorius Luyten, None
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science June 2015, Vol.56, 2957. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Jan-Willem M Beenakker, Denis P Shamonin, Andrew G Webb, Berend C Stoel, Gregorius P M Luyten; Improved 3D retinal shape models with high-field MRI. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):2957.

      Download citation file:


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

      ×
  • Supplements
Abstract
 
Purpose
 

To develop patient-specific 3D models of the retina for ray-tracing simulations.<br /> Conventional optical techniques used to measure distances in the eye have significant systematic errors for off-axis measurements due to refraction, which “bends” the light beams. Since MRI is not affected by refraction it is an important tool to study the retinal shape. However, the sensitivity of MRI to eye-motion has limited these evaluations to 2D or low-resolution 3D data. In this context, we describe a method that uses the advances in high field MRI to quantify the full 3D retinal shape with high accuracy.

 
Methods
 

We examined 11 emmetropic (|refractive error| < 0.5D) and 10 myopic (refractive error between -0.5D and -7D) subjects with no further ocular pathologies. The examination consisted of an ocular MRI scan and an auto-refraction measurement. 4 subjects have been scanned twice to assess the methods reproducibility.<br /> The MRI was performed on a Philips 7 Tesla MRI. The left eye was scanned with a dedicated 3-channel receive eye-coil. Eye-motion artefacts were minimized with a cued-blinking protocol[Beenakker, NMR Biomed 2013]. The MRI examination takes less than 15 minutes.<br /> The resulting MR-images were segmented automatically with sub-pixel precision. The retinal shape was quantified by fitting an ellipsoid to the detected contours both in 3D and in 2D cross-sections. The accuracy and reproducibility of each fit were recorded for a range of percentages of included points.

 
Results
 

Fig. 1 gives an example of a 2D ellipse fit to the retina. The 2D fits converge to similar ellipses over a wide range of number of points used, fig. 2. The resulting shape descriptors, such as retinal curvature, are highly reproducible when more than 30% of the contour is used. The 3D ellipsoid fits reproduce previous findings such as the average increase of retinal curvature in emmetropia compared to myopia (74.8m-1 vs 77.0m-1).<br /> Furthermore, the full 3D assessment shows that most eyes (17/21) have an oblique orientation, which cannot be measured with conventional 2D scans.

 
Conclusions
 

Full 3D MR-images of the eye allow for a better description of the retinal shape, which will advance the design of patient-specific eye-models and has potential applications in the study of myopia development.  

 
Fig 1: Elliptic fit on high-resolution MR-images
 
Fig 1: Elliptic fit on high-resolution MR-images
 
 
Fig 2.: Goodness of fit of all subjects and horizontal retinal curvature of one subject as a function of percentage of included points.
 
Fig 2.: Goodness of fit of all subjects and horizontal retinal curvature of one subject as a function of percentage of included points.

 
×
×

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.

×