Abstract
Purpose :
Magnetic resonance images (MRI) of the eye and optical coherence tomography (OCT) images of the posterior pole have demonstrated morphological changes in the structure of the eye in association with high intracranial pressure (ICP). Patient-specific numerical models have the potential to advance understanding regarding on the mechanical basis for these changes. Methods to develop patient specific models of eye geometry that serve as the foundation for mechanical models have been reported based on histology of the optic nerve complex for glaucoma applications and MRI of the globe for myopia studies. We sought to develop methodology for creating patient specific models of eye geometry with a focus on the optic nerve complex based on MRI of human eyes in high ICP states.
Methods :
An axial slice (Fig 1) that captured a mid-globe cross was selected from 3D Fast Imaging with Steady State Acquisition (3D FIESTA-C) MR images (0.625 mm x 0.625 mm resolution) that were clinically acquired on a 3T MRI device (GE Healthcare) prior to ICP treatment in a patient with high ICP and papilledema. Threshold paint tools were used to mask the sclera, vitreous, optic nerve, subarachnoid space and optic nerve sheath. These were meshed with 2-D triangle elements (Scan IP, Synopsys Inc) (Fig 2).
Results :
The resulting 2D eye geometry was comprised of 23548 elements representing sclera, 7783 elements representing optic nerve and 14070 elements representing optic nerve sheath. Element apices (nodes) are shared between adjacent tissues to ensure coupled deformation. The cartesian coordinates defining nodes, their combination to define elements and sets of nodes that define surfaces can be exported for use in creation of numerical models of eye biomechanics using finite element analysis (FEA).
Conclusions :
We report a method for developing patient specific models of eye geometry in high ICP states based on MRI scans. These can serve as the foundation for FEA models of ocular biomechanics and possibly a source for biomarkers of ICP. Future directions include expansion of the methodology to create 3D models and combining multiple imaging modalities to achieve more detailed representation of optic nerve head structures.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.