Retinal layer segmentation in patients with multiple sclerosis using spectral domain optical coherence tomography.

2014 
Purpose To evaluate the thickness of the 10 retinal layers in the paramacular area of patients with multiple sclerosis (MS) compared with healthy subjects using the new segmentation technology of spectral domain optical coherence tomography (OCT). To examine which layer has better sensitivity for detecting neurodegeneration in patients with MS. Design Observational, cross-sectional study. Participants Patients with MS (n = 204) and age-matched healthy subjects (n = 138). Methods The Spectralis OCT system (Heidelberg Engineering, Inc., Heidelberg, Germany) was used to obtain automated segmentation of all retinal layers in a parafoveal scan in 1 randomly selected eye of each participant, using the new segmentation application prototype. Main Outcome Measures The thicknesses of 512 parafoveal points in the 10 retinal layers were obtained in each eye, and the mean thickness of each layer was calculated and compared between patients with MS and healthy subjects. The analysis was repeated, comparing patients with MS with and without previous optic neuritis. Correlation analysis was performed to evaluate the association between each retinal layer mean thickness, duration of disease, and functional disability in patients with MS. A logistic regression analysis was performed to determine which layer provided better sensitivity for detecting neurodegeneration in patients with MS. Results All retinal layers, except the inner limiting membrane, were thinner in patients with MS compared with healthy subjects ( P P Conclusions Analysis based on the segmentation technology of the Spectralis OCT revealed retinal layer atrophy in patients with MS, especially of the inner layers. Reduction of the ganglion cell and inner plexiform layers predicted greater axonal damage in patients with MS.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    25
    References
    77
    Citations
    NaN
    KQI
    []