Purchase this article with an account.
B. Gilmartin, N. S. Logan, K. D. Singh; Determining Regional Variations in Globe Conformation Using 3-D Ocular MRI. Invest. Ophthalmol. Vis. Sci. 2007;48(13):1215. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To demonstrate novel structural features of the eye globe using a new technique that has characterised non-invasively the 3-D shape of the human eye using Magnetic Resonance Imaging (MRI) (Singh et al. IOVS, 2006, 47, 2272-2279).
Twenty participants, selected to have a range of iso- and anisometropic errors were scanned using a Siemens Trio 3-tesla whole-body MR scanner. The technique involves the acquisition of a T2-weighted MRI which is optimised to reveal the fluid-filled chambers of the eye. For analysis of MR images the eye is initially enveloped by a sphere which comprises 32,768 3-D triangular polygons of equal area distributed uniformly across its surface. Automatic segmentation and meshing algorithms generate a 3-D surface model by an iterative shrink wrap process centred about the geometrical centre of the initial sphere. The process alters the position of the vertices of each polygon which results in the redistribution and resizing of polygons across the surface of the eye model. The 3D vector co-ordinates of each of the modified polygons allows them to be assigned to one of a series of annular segments each of which map to successive increments along the longitudinal axis equal to 1% of the total axial length (h). The area (A) of triangles falling within each successive segment are summed and converted to an equivalent spherical radius (r) using the relationship r = A/2πh. The variation in r is plotted across the 2nd, 3rd and 4th quartiles of axial length; the incremental change in r thus providing an index of change in globe conformation.
Data for r were shown to be repeatable for 1% increments between 25% and 95% of axial length (repeated measures ANOVA for 9 separate runs on an individual subject p>0.05). All subjects exhibited a spherical globe conformation across the 2nd and 3rd quartiles. Although two subjects retained sphericity across the subsequent 75-95% of the 4th quartile, globe conformation for the remaining subjects exhibited significant steepening or flattening which was independent of isometropic category but showed inter-eye differences in anisometropic subjects.
We speculate that the spherical constancy present in the 2nd and 3rd quartiles results from pre-programmed growth whereas the structural variance in the 4th quartile results from vision-dependent modulated growth. Of note is that the angular subtense of the 4th quartile approximates to the binocular overlapping visual field when coupled with the fellow eye.
This PDF is available to Subscribers Only