Elsevier

Ophthalmology

Volume 114, Issue 2, February 2007, Pages 263-270
Ophthalmology

Original Article
Alignment Artifacts in Optical Coherence Tomography Analyzed Images

Poster presented at: American Academy of Ophthalmology annual meeting, October 2005, Chicago, Illinois.
https://doi.org/10.1016/j.ophtha.2006.06.059Get rights and content

Purpose

To compare retinal thickness and retinal height of the original scanned optical coherence tomography (OCT) images with those of the same images after automated retinal thickness analysis.

Design

Cross-sectional study.

Participants

Thirty normal eyes, 20 eyes with neovascular age-related macular degeneration (AMD), 20 with central serous chorioretinopathy (CSC), 20 with macular holes, and 20 with non–AMD related macular edema from 110 subjects were selected randomly from an OCT database. One of the 6 macular scans in each eye was chosen randomly for analysis.

Methods

Two sets of OCT images—original images and analyzed images (after retinal thickness [single eye] analysis)—in each eye were exported for retinal thickness or retinal height measurement. Comparisons of retinal thickness or retinal height at selected locations were performed.

Main Outcome Measures

Retinal thickness (defined as the distance between the anterior boundary of the retinal nerve fiber layer [RNFL] and posterior boundary of the photoreceptor layer) and retinal height (defined as the distance between the anterior boundary of the RNFL and baseline level of the anterior boundary of the retinal pigment epithelium [RPE]). Retinal height was measured when there was detachment of neurosensory retina or RPE.

Results

No significant difference in retinal thickness was observed between the original and the analyzed OCT images in normal eyes and in eyes with macular holes or non–AMD related macular edema. However, OCT-analyzed images demonstrated retinal thickness or retinal height measurements in eyes with CSC or neovascular AMD significantly lower than the corresponding measurements in the original images (all with Ps≤0.001, Wilcoxon signed rank test). In the groups of neovascular AMD and CSC, Bland–Altman plots revealed mean differences of 124 μm (95% limits of agreement between −65.5 and 313.6) and 84.4 μm (95% limits of agreement between −178.0 and 346.8), respectively, between the original and analyzed retinal measurements.

Conclusions

Retinal thickness and retinal height could be underestimated in patients with CSC or neovascular AMD after retinal thickness analysis in Stratus OCT when either automatic measurements or manual caliper–assisted measurements are performed on the analyzed images. We recommend exporting the original scanned OCT images for retinal thickness and retinal height measurement in patients with CSC or neovascular AMD.

Section snippets

Materials and Methods

The study was conducted in accordance with the ethical standards stated in the Declaration of Helsinki. Approval was obtained from the Clinical Research Ethics Committee of Hong Kong Hospital Authority Kowloon West Cluster to review the medical records of individuals who received OCT scans over the macular region during the period from July 2003 to June 2005 in the Department of Ophthalmology, Caritas Medical Centre, Hong Kong. All the OCT images were obtained using Stratus OCT with the use of

Comparisons between Manual and Automated Retinal Thickness Measurements in Normal Eyes

Manually measured foveal thicknesses from 30 normal eyes were compared with those generated automatically by OCT. After retinal thickness (single eye) analysis, 1 of the 6 linear macular images was selected randomly and exported to SigmaScan Pro. The foveal thickness at the 65th A-scan was measured (as described in “Materials and Methods”) and compared to the automated measurements at the corresponding position in each eye (Fig 3). The mean/median (range) foveal thicknesses were 149.7/147

Discussion

Reliable and accurate delineation of retinal boundaries for measurement of macular thickness in OCT macular scans is crucial in detecting macular abnormality, monitoring disease progression, and assessing therapeutic response. However, even with the latest commercially available OCT unit (Stratus OCT) and its attached image analysis software (version 4.0), retinal thickness miscalculation is still a commonly encountered problem. The major causes of the inaccuracies in retinal thickness

References (17)

There are more references available in the full text version of this article.

Cited by (0)

Manuscript no. 2005-843.

The authors have no proprietary interest in the development or marketing of any product mentioned in the article, and the study received no financial support.

View full text