Purchase this article with an account.
Leonard Seibold, Naresh Mandava, Malik Y. Kahook; Retinal Nerve Fiber Layer Thickness Measurements after Optical Coherence Tomography Software Upgrade. Invest. Ophthalmol. Vis. Sci. 2011;52(14):3676.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To determine the impact of software upgrades on retinal nerve fiber layer thickness (RNFL) measurements taken by spectral domain optical coherence tomography (OCT).
80 normal eyes (40 patients) were scanned for RNFL thickness measurements using the Spectralis (Heidelberg Engineering, Heidelberg, Germany) spectral-domain OCT instrument by a single experienced operator on the same day. All eyes underwent initial baseline examination including Snellen visual acuity, intraocular pressure measurement, gonioscopy, optic nerve head exam, and Humphrey visual field examination. Scan analysis was performed using version 4.0 software and then raw data reanalyzed after upgrade to version 5.1.3. Student paired t testing and Pearson's correlation coefficient were used for statistical analysis.
The correlation of average and quadrant RNFL thicknesses generated using version 4.0 and 5.1.3 software on Spectralis was very high. For average RNFL thickness, Pearson correlation coefficient was 0.998 while quadrant thickness ranged 0.996-0.957. The average RNFL thickness using version 4.0 software was 107.41 µm ± 13.30 and 107.34 µm ± 13.13 for version 5.1.3 software. The difference of +0.07 µm did not reach significance (p=0.488). Version 5.1.3 also generated thinner RNFL measurements compared to its predecessor in the temporal, nasal, and inferior quadrants. Mean quadrant RNFL differences between version 4.0 and 5.1.3 were +0.37 µm (p=0.055) temporal, -1.15 µm (p=0.096) superior, +0.26 µm (p= 0.294) nasal, and 0.00 µm (p=1.000) inferior.
Recent studies have shown significant differences in RNFL measurements from one OCT manufacturer to the next and from time-domain to spectral-domain. This is the first study to evaluate for differences in thickness values due solely to upgrades in software version. There is excellent correlation in RNFL thickness measurements taken on Spectralis OCT and analyzed before and after software upgrades. While the thickness values are not identical, the difference did not reach statistical significance. Clinicians should be aware of these relationships when upgrading software versions in practice.
This PDF is available to Subscribers Only