Purchase this article with an account.
JL Hougaard, B Sander; Nerve Fiber Layer Thickness as Measured by Optical Coherence Tomography: A Comparative Study Between the Software Algorithm Version A4.1 and A6.1 . Invest. Ophthalmol. Vis. Sci. 2002;43(13):271.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: To compare nerve fiber layer thickness measurements using two different commercially available optical coherence tomography software algorithms. Methods: In 60 eyes (35 persons), 3 circular peripapillary optical coherence tomography scans per eye were analysed regarding nerve fiber layer thickness (NFLT) using the software algorithm version A4.1 (windows 3.11) and version A6.1 (windows 98) after a system upgrade. Persons were included if algorithm problems did not occur in detecting the nerve fiber layer in any single scan using any of the two algorithms. The comparison was performed on NFLT as a mean of the three scans per eye. The scans compared varied over a wide range of NFLT values as scans originated from patients with glaucoma, glaucoma suspects (including intraocular hypertensive), and healthy eyes. Results: On average (n = 60) the full circle NFLT was 75.4 27.8 microns ( 1 SD) using A4.1 vs. 111.3 26.8 microns using A6.1, the mean difference being 35.9 6.6 microns ( SD) which was significant (p < 0.0001). A high correlation between algorithms was found (r = 0.97, p < 0.001). For full circle NFLT the difference between the algorithms was not correlated to the measured NFLT value (r = -0.15, p = 0.25). However, for superior and inferior quadrant NFLT means, respectively, the difference between algorithms showed a significant correlation to the measured quadrant NFLT (r = -0.34, p = 0.0078; r = -0.49, p < 0.0001). Conclusion: The software algorithm version A4.1 and A6.1 correlate but do not agree on average measuring the NFLT. It seems that over a wide range of full circle NFLT values the mean difference was not correlated to the measured full circle NFLT, but for the regions of specific interest (superior and inferior quadrants) the difference between algorithms was pronounced for the very thin NFLT. This finding indicate that algorithms measuring the NFLT using optical coherence tomography is an issue for further investigation.
This PDF is available to Subscribers Only