Purchase this article with an account.
Bhavna J Antony, Byung-Jin Kim, Donald J Zack, Peter A. Calabresi, Jerry L. Prince; Characterization of Retinal Layer Thickness Changes Using Volumetric SD-OCT Images in a Multiple Sclerosis Mouse Model. Invest. Ophthalmol. Vis. Sci. 2016;57(12):2199.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To characterize longitudinal retinal layer thickness changes in the experimental autoimmune encephalomyelitis (EAE) mouse model using SDOCT in order to validate this imaging modality as an outcome measure of neurodegeneration and for examining the efficacy of putative neuroprotective drugs.
Ten female C57BL/6J mice (8 weeks) were immunized with MOG 35-55 (100 μg) for EAE induction. SD-OCT volumetric images were acquired from both eyes using a Bioptigen SD-OCT scanner where the volumes were obtained within a 1.4mm x 1.4mm x 1.53mm region. Eight retinal layers were segmented using an automated graph-theoretic approach, initially proposed for human scan analysis and adapted for murine scans (see Fig. 1). This method uses a random forest to learn surface descriptors from an independent training set. The average thickness was computed for the retinal layers at each time point and the longitudinal changes were tracked. The baseline and day 56 average thicknesses were statistically compared using a student’s t-test. Disease severity was used to divide the mice into 2 groups, GP1 and GP2, where GP1 showed mild/no disease. The coefficient of variation (CV) at each time point (with respect to the baseline thickness) for the retinal layer thicknesses were also computed for GP 1.
The NFGC+IPL (see Fig. 1), initially 63.82 +/- 0.85 μm, thinned significantly in 56 days in GP2 of mice to 56.59 +/- 2.42 μm (p < 0.01), while the average thickness in GP1 was 64.39 +/- 2.58 μm. The total retinal thickness (TRT) also changed significantly, reducing from 184.97 μm at baseline to 178.61 +/- 3.94 μm at day 56 in GP2 (p < 0.01) but was 186.12 +/- 4.26 μm in GP1. The NFGC and the inner nuclear layer thinned marginally (p < 0.05). No changes were noted in the outer nuclear layer or in the outer retina. The CV for all the retinal layers in GP1 at the 8 time points were less than 5%, with the largest CV (4.43%) being noted in the NFGC.
The automated analysis of SD-OCT volumes is a powerful tool to monitor progressive retinal changes in the EAE mouse model, indicated by the consistently low CV numbers. The inner plexiform layer in particular, thinned quite dramatically over the 8-week study. Furthermore, OCT may be a useful serial in vivo measure of retinal neuron pathology in this model.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
Section of Bscan, showing segmented retinal layers
Retinal layer thickness tracjectory
This PDF is available to Subscribers Only