Purchase this article with an account.
M. Wendt, K. H. Bockhorst, L. He, A. Glasser; How Accurately Can Lens Volume Be Measured With Magnetic Resonance Imaging?. Invest. Ophthalmol. Vis. Sci. 2010;51(13):795.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Magnetic Resonance Imaging (MRI) has been used to image the accommodative structures in the living human eye. However, MR imaging has limited spatial and temporal resolution. This study was undertaken to determine how accurately lens volume can be measured from MRI images.
Nineteen eyes from ~6 month old pigs were inflated to 15 mm Hg. Eyes were glued in a saline filled tube. MRI scans were performed with a 7 Tesla Biospec (Bruker, Karlsruhe, Germany) using a linear volume transmitter coil and a quadrature surface receiver coil. 2D Rapid Acquisition and Relaxation Enhancement (RARE factor 32, TR 18 s, TE 13 ms) scans of 14 minutes were performed. Slice thickness was 0.5 mm with a 32 x 32 mm field of view, a matrix of 256 x 256, and a spatial resolution of 0.125 x 0.125 x 0.5 mm. Eyes were then dissected to remove the lens and the lens photographed in saline. Isolated lenses were weighed and lens volume was measured using a fluid displacement method. Lens volumes were calculated from the central MRI slice and biometry images respectively and from the entire MRI stack using edge detection of images transformed to spherical coordinates.
Lens volume from biometry images related to actual measured lens volume by y = 0.9404x + 0.0046 (r2 = 0.8231). Lens volume calculated from the central MRI slice related to actual volume by y = 1.1182x - 0.0228 (r2 = 0.9644). Lens volume calculated from the MRI stack related to actual measured volume by y = 0.9498x + 0.0242 (r2 = 0.9628).
Volumes calculated from all three methods showed significant linear relationships which were similar to the ideal 1:1 lines (p = 0.28 for MRI stack, p = 0.59 for biometry, p = 0.058 for MRI slice). However, volumes calculated from a single MRI slice and from a single image of the isolated lens had greater variability than those calculated using whole MRI stacks. These near ideal MR imaging methods using enucleated eyes permit lens volumes to be measured to an accuracy of 9.04 mm3 (MRI stack), 25.25 mm3 (MRI slice), and 24.36 mm3 (biometry). In vivo MR imaging in living human eyes would result in reduced resolution and accuracy due to eye movements, blinks and the need for shorter imaging time.
This PDF is available to Subscribers Only