Article Text

Download PDFPDF

Volumetric ellipsoid zone mapping for enhanced visualisation of outer retinal integrity with optical coherence tomography
  1. Yuji Itoh1,
  2. Amit Vasanji2,
  3. Justis P Ehlers1
  1. 1Ophthalmic Imaging Center, Cole Eye Institute, Cleveland Clinic, Cleveland, Ohio, USA
  2. 2ImageIQ, Cleveland, Ohio, USA
  1. Correspondence to Dr Justis P Ehlers, Cole Eye Institute, Cleveland Clinic, 9500 Euclid Avenue, i30, Cleveland, OH 44195, USA; ehlersj1{at}yahoo.com

Abstract

Objective assessment of retinal layer integrity with optical coherence tomography (OCT) is currently limited. The ellipsoid zone (EZ) has been identified as an important feature on OCT that has critical prognostic value in macular disorders. In this report, we describe a novel assessment tool for EZ integrity that provides visual and quantitative assessment across an OCT data set. Using this algorithm, we describe the findings in multiple clinical examples, including normal controls, age-related macular degeneration, drug effects (eg, ocriplasmin, hydroxychloroquine) and effects of surgical manipulation (eg, following membrane peeling using intraoperative OCT). EZ mapping provides both en face visualisation of EZ integrity and EZ-retinal pigment epithelium height. Additionally, volumetric, area and linear measurements are feasible using this assessment tool.

  • Retina
  • Imaging
  • Macula
  • Pathology

Statistics from Altmetric.com

Introduction

In the past decade, optical coherence tomography (OCT) has transformed the diagnostic approach to numerous vitreoretinal diseases. Spectral-domain OCT (SDOCT) has allowed for outstanding visualisation of numerous retinal structures. The outer retina is characterised by four highly reflective bands that can be seen on SDOCT in the normal eye (ie, external limiting membrane (ELM), ellipsoid zone (EZ), interdigitation zone (IZ, or cone outer segment tips) and retinal pigment epithelium (RPE)).1–3 The EZ and ELM, in particular, have been linked to visual outcomes and prognosis in numerous macular conditions, such as age-related macular degeneration (AMD), hydroxychloroquine toxicity and intravitreal ocriplasmin injection.4–10 Objective evaluation of the EZ has been lacking. The purpose of this report was to evaluate a novel analysis tool for EZ mapping with en face visualisation and volumetric assessment in numerous retinal conditions as a potential platform for potential future research and clinical applications.

Methods

Cleveland Clinic IRB approval was obtained for a retrospective assessment of eyes undergoing SDOCT testing with a novel EZ mapping tool. Initial conditions selected for evaluation included normal controls, geographic atrophy secondary to AMD, hydroxychloroquine toxicity, ocriplasmin-related EZ attenuation and outer retinal dynamics following membrane peeling. These conditions were selected due to the predominance of outer retinal alterations. Additionally, for initial assessment purposes, conditions with minimal RPE disturbance were selected.

An automated EZ mapping tool was developed at the Ophthalmic Imaging Center at Cleveland Clinic through collaboration with ImageIQ (Cleveland, Ohio, USA) for segmenting the EZ with additional retinal layers, providing linear, area and volumetric measurements, as well as en face visualisation for evaluating EZ and outer retinal dynamics (figure 1). In brief, the macular cube data set was imported into a novel OCT-automated segmentation tool. Automated segmentation and mapping were performed. The SDOCT image stack was passed through multiple filters and the RPE was identified as the floor for the algorithm segmentation. Using a prespecified search area, the EZ was identified, if present, based on relative intensity, proximity and relative location of the outer retinal bands. Manual independent reviewer of the segmentation accuracy was performed for each of the eyes. To evaluate reliability of the algorithm, multiple time points in normal samples were assessed and compared for variability. The data set was then transformed into multiple output components. Cross-sectional area and cubic volumetric data were generated related to the EZ–RPE thickness. Three-dimensional reconstruction was also performed of the entire data set with embedded EZ mapping for visualisation of areas of pathology. Finally, en face EZ thickness topographic maps were created for each data set. Zeiss Cirrus SDOCT system was used for all normal, AMD, hydroxychloroquine and ocriplasmin eyes with a 6 mm×6 mm macular cube. The intraoperative OCT scan assessment following membrane peeling was performed with the Bioptigen Envisu system with a 10 mm×10 mm macular cube.

Figure 1

Ellipsoid zone mapping in a normal eye. (A) B-scan segmentation of the ellipsoid zone. (B) Three-dimensional reconstruction of the ellipsoid zone revealing the overall integrity (green). (C) En face mapping of relative ellipsoid zone to retinal pigment epithelium thickness.

Results

A total of 34 eyes were included in this innovation report. Twelve normal eyes were assessed with the EZ mapping algorithm. The mean EZ–RPE volume was 1.27±0.094 (mm3), the mean central foveal area was 0.22±0.019 (mm2) and mean EZ map thickness was >20 µ in 99% of sampled areas. Multiple time points were assessed in 11 normal eyes to evaluate the algorithm reproducibility. The mean EZ volume at the first and second visits were highly correlated (1.26±0.094 mm3 and 1.28±0.11 mm3, correlation coefficient=0.82, p<0.001).

EZ mapping in AMD with geographic atrophy (n=4) exhibited multifocal areas of EZ loss on the en face map and decreased EZ–RPE volume (figure 2). EZ–RPE volume was reduced to a mean of 1.02±0.17 mm3 (p=0.015, compared with normal cases) and an EZ thickness of >20 µ was seen in only (59%–89%) of the cube compared with 99% in normal eyes reflecting the areas of atrophy.

Figure 2

Ellipsoid zone mapping in an eye with geographic atrophy secondary to age-related macular degeneration. (A) B-scan segmentation of the ellipsoid zone. (B) Three-dimensional reconstruction of the ellipsoid zone revealing the overall integrity (green) and areas of ellipsoid zone loss (blue). (C) En face mapping of relative ellipsoid zone to retinal pigment epithelium thickness, note the reduced ellipsoid zone to retinal pigment epithelium thickness (arrows).

Mild hydroxychloroquine toxicity (n=3) EZ mapping exhibited the expected concentric thinning of the EZ–RPE thickness with associated foveal sparing. In a representative case (figure 3), EZ–RPE volume and area were reduced compared with normals (0.93 and 0.17 mm2, respectively). The EZ map showed significant reduction EZ integrity with a mean EZ thickness of >20 µ in only 80% of sampled areas.

Figure 3

Ellipsoid zone mapping in an eye with early hydroxychloroquine toxicity. (A) B-scan segmentation of the ellipsoid zone. (B) Three-dimensional reconstruction of the ellipsoid zone revealing the overall integrity (green) and areas of annular ellipsoid zone loss (blue). (C) En face mapping of relative ellipsoid zone to retinal pigment epithelium thickness, note the reduced ellipsoid zone to retinal pigment epithelium thickness (arrows).

Ocriplasmin-related outer retinal attenuation (n=15) showed variable EZ attenuation and thinning (figure 4). In a representative case, EZ mapping revealed reduction of EZ–RPE volume to 0.15 mm3. EZ thickness of >20 µ was present in only 52% of sampled areas on EZ en face maps.

Figure 4

Ellipsoid zone mapping in an eye prior to and after ocriplasmin injection. (A) B-scan segmentation of the ellipsoid zone prior to ocriplasmin injection. (B and C) Three-dimensional reconstruction of the ellipsoid zone revealing the overall ellipsoid zone integrity (green) and automated segmentation/reconstruction of macular hole (red). (D) En face mapping of relative ellipsoid zone to retinal pigment epithelium thickness, note the absent ellipsoid zone in area of macular hole. (E) B-scan segmentation of the ellipsoid zone following ocriplasmin injection. Note significant increase in subretinal and macular hole size (yellow). (F and G) Three-dimensional reconstruction of the ellipsoid zone revealing the overall ellipsoid zone integrity (green) and multifocal areas of ellipsoid zone loss (blue). Macular hole enlargement is also identified (red). (H) En face mapping of relative ellipsoid zone to retinal pigment epithelium thickness, note the expansion ellipsoid zone loss in area of enlarged macular hole and the diffuse heterogenous thinning of the ellipsoid zone.

EZ mapping was also performed on intraoperative OCT scans before and after internal limiting membrane (ILM) peeling. Prior to ILM peeling the EZ–RPE volume was 2.94 mm3 and after ILM was increased dramatically to 4.7 mm3. This was reflected in the en face EZ map (figure 5).

Figure 5

Ellipsoid zone mapping with intraoperative optical coherence tomography scans before and after internal limiting membrane (ILM) peeling in a macular hole case. (A) B-scan prior to ILM peeling showing typical ellipsoid zone to retinal pigment epithelium height. (B) Three-dimensional reconstruction of the ellipsoid zone revealing the overall integrity (green) and segmented macular hole (red). (C) En face mapping of relative ellipsoid zone to retinal pigment epithelium thickness. (D) Following ILM peeling, B-scan reveals expansion of the ellipsoid zone to retinal pigment epithelium height. (E) Three-dimensional reconstruction of the ellipsoid zone revealing the overall maintained integrity (green) and segmented macular hole (red). (F) En face mapping of relative ellipsoid zone to retinal pigment epithelium thickness, note marked diffuse increased ellipsoid zone to retinal pigment epithelium height.

Discussion

This study represents a new innovation in OCT assessment of relative retinal layer analysis with particular focus on the EZ and outer retina. In-depth visualisation using EZ mapping allows for overall volumetric assessment and topographic visualisation of the relative locations of EZ alterations. In this report, we use the novel tool for evaluation of numerous macular conditions.

In normal eyes, the EZ maps revealed excellent EZ integrity with high repeatability. In eyes after intravitreal ocriplasmin, geographic atrophy and hydroxychloroquine toxicity, EZ mapping successfully identified EZ–RPE alterations. Not only could this be assessed on a volumetric scale, but topographic representations of the en face EZ map allowed for pattern-based analysis and identification. This may be particularly useful in subtle bulls-eye patterns (eg, subclinical hydroxychloroquine toxicity) and for multifocal extrafoveal atrophy, such as in early geographic atrophy. Change analysis of thickness maps may also be particularly useful in understanding retinal dynamics and alterations that occur during various disease processes or following therapeutic interventions. For example, alterations in the EZ–RPE relationship following ILM peeling identified on intraoperative OCT predicted speed of anatomic normalisation in macular hole repair.11 Using the EZ mapping system, rapid quantitative assessment of intraoperative alterations may be feasible that prognosticates macular hole normalisation rate.

This report has some important limitations, including its small sample size and its retrospective nature. Further prospective research with larger numbers including normal eye, various retinal diseases are needed to assemble a normative database. However, the utility of an objective quantitative assessment tool for EZ integrity for clinical trials and disease prognostication/management may prove particularly useful. This report was primarily limited to conditions that more focally impact the outer retina with a minimally disturbed RPE. Expansion of the assessment tool is needed into more complex pathologies, such as choroidal neovascularisation and conditions associated with macular oedema (eg, diabetic macular oedema). This research is ongoing.

This report describes a novel technology and visualisation tool for the quantitative assessment of the EZ and EZ–RPE thickness using both volumetric and en face tomographic representations. Additional research is needed to further assess the utility of the tool in various pathological conditions, as well as more in-depth integrative pattern analysis is needed to better assess this tool as a disease modelling and diagnosis guidance tool.

References

Footnotes

  • Contributors The following outlines the contributions of the authors: design of the study (JPE); conduct of the study (all authors); data collection (JPE, YI); data management (JPE, YI); data analysis (all authors); data interpretation (all authors); preparation of the manuscript (all authors); review and approval of the manuscript (all authors).

  • Funding NIH/NEI K23-EY022947-01A1 (JPE); Ohio Department of Development TECH-13-059 (JPE); Research to Prevent Blindness (Cole Eye Institutional Grant); Machemer Foundation Fellowship (YI).

  • Competing interests JPE: Bioptigen (C, P), Thrombogenics (C, R), Synergetics (P), Genentech (R), Leica (C), Carl Zeiss Meditec (C) and Alcon (C).

  • Ethics approval Cleveland Clinic IRB.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Analysis tool will be made publicly available when final platform finalised to user-friendly interface.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.