The recent article by Vizzeri and associates makes an interesting contribution to the emerging literature examining the roles of signal strength and scan misalignment on the reliability of retinal nerve fibre layer (RNFL) measurements using optical coherence tomography (OCT). In this retrospective study, the authors assessed the relationship between signal strength and RNFL measurements by comparing repeat and baseline Stratus OCT (Carl Zeiss Meditec, Dublin, California, USA) scans. Influences from vertical and horizontal scan alignment on RNFL measurements also was assessed. The authors reported a larger variation in RNFL measurements between scans at different visits than on repeat scans at the same visit, and also reported significant differences (from baseline) in quadrant-specific RNFL measurements with signal strength. A positive association between signal strength and average RNFL measurements at signal strength 7 or more was noted. Also, signal strength influenced the nasal minus temporal RNFL quadrant measurements but not superior minus inferior RNFL quadrant measurements, suggesting that RNFL measurement variability was associated more strongly with horizontal than with vertical misalignment.
We assessed RNFL measurements in 2092 healthy adolescents aged predominantly 12 years using Stratus OCT as part of the Sydney Myopia Study. Signal strength was classified into 3 groups of incremental quality. Moderate scan quality was classified as signal strengths of 5, 6, or 7; good scan quality had signal strengths of 8 or 9; and excellent scan quality had a signal strength of 10. In contrast to the authors’ findings, we could not document any association between average RNFL thickness and signal strength category, regardless of age, gender, ethnicity, axial length, and spherical equivalent refraction. However, a positive association was found in the temporal quadrant between RNFL thickness and signal strength, with excellent scans producing significantly thicker RNFL measurements than both moderate and good scans.
The influence of signal strength on OCT measurements has been explored previously with mixed results. In a glaucomatous population, Asrani and associates reported 4% of Stratus OCT RNFL measurements of signal strength 6 or more had artefacts, whereas 86% of measurements had artefacts when signal strength dropped to 4 or less. This suggests that low signal strength may increase the presence of artefacts and that it influences measurement accuracy. In addition, in assessing 165 subjects with glaucoma or optic atrophy, Wu and associates found a positive association between signal strength and RNFL thickness measurements between signal strengths of 5 and 7. However, they found no statistical differences in RNFL measurements with signal strengths of 7 or more. In contrast, Cheung and associates assessed a sample of 40 nonglaucomatous adult subjects using Stratus OCT and found that RNFL measurements were significantly thicker with signal strength of 10 compared with lower signal strengths.
Thus, despite a likely association between signal strength and RNFL thickness measurements, it seems that a definitive signal strength cutoff for acceptable quality in Stratus OCT measurements has yet to be identified. Pooling data from recent studies may help to clarify this important clinical question. It would also be interesting to assess whether signal strength, also used in the newer spectral-domain OCT, is associated with differences in RNFL thickness measurements.