Purchase this article with an account.
Arezoo Miraftabi, Navid Amini, Sharon Henry, Simon K Law, Anne L Coleman, Joseph Caprioli, Kouros Nouri-Mahdavi; Regional Intra-session variability of full and sublayer macular thickness measurements with SD-OCT. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):4531.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare the intra-session variability of ganglion cell (GCL), inner plexiform layer ( IPL), and full macular thickness measurements at superpixel level in glaucoma and normal subjects.
102 patients and 21 normal subjects from the Advanced Glaucoma Progression Study who had three macular imaging by one operator in the same visit were enrolled.The OCT scans (Posterior Pole algorithm) were performed with the Spectralis SD-OCT and retinal layer were segmented by the GMPE(The Glaucoma Module Premium Edition) software.Mean and SD values and measurement errors (difference between each of the triplicate values and the mean of triplicate values) were calculated in each cell (super pixel from the 8x8 array) for all outcome measures. Measurement error correlation across locations for a given layer and group were computed using a components of variance model.Correlation between variability and thickness measurements was compared among normal and glaucoma groups.
The mean (±SD) age was 66.6 (±9.6) years in the glaucoma group and 60.5(±9.0) years in the normal subjects. Visual field mean deviation was −8.7 (±5.5) dB and −0.4(±0.9) dB in glaucoma and normal subjects, respectively. Prediction bounds for the measurement error was 4.0 (±1.0) μm for IPL, 4.0 (±1.1) μm for GCL and 6.3 (± 2.2) μm for full macular thickness in the glaucoma group vs. 3.3(±0.8) μm ,3.3(± 0.8) μm ,and 4.7(±1.2) μm in normal subjects.Variances showed a small but significant difference between glaucoma and normal subjects (p<0.001). Measurement error was random across the 64 superpixels although more variability was seen in the nasal area. Variability was mostly independent of mean thickness although there was a tendency for GC measurement errors to increase with decreasing thickness in glaucoma patients (Figure).The variabilty of superpixels around fovea were not significantly different compared to other areas (p>0.05).Prediction bounds were uniform across eyes and subjects.
Variability of regional macular full thickness and sublayer measurements is small and consistent within and across eyes confirming potential utility of such measurements in glaucoma patients.Glaucoma eyes tend to have slightly larger variability compared to normal eyes. Given the results,defining a variability space for individual eyes seems unnecessary and therefore , taking only one SD-OCT image at baseline seems to be adequate.
This PDF is available to Subscribers Only