Article Text

other Versions

Download PDFPDF
Comparison of retinal thickness measurements and segmentation performance of four different spectral and time domain OCT devices in neovascular age-related macular degeneration
  1. Georgos Mylonas,
  2. Christian Ahlers,
  3. Panagiotis Malamos,
  4. Isabel Golbaz,
  5. Gabor Deak,
  6. Christopher Schütze,
  7. Stefan Sacu (stefan.sacu{at},
  8. Ursula Schmidt-Erfurth
  1. Medical University of Vienna, Austria
  2. Medical University of Vienna, Austria
  3. Medical University of Vienna, Austria
  4. Medical University of Vienna, Austria
  5. Medical University of Vienna, Austria
  6. Medical University of Vienna, Austria
  7. Medical University of Vienna, Austria
  8. Medical University of Vienna, Austria


    Aims: To evaluate the reliability of different optical coherence tomography (OCT) devices and scanning patterns in the assessment of retinal thickness and segmentation performance in neovascular age related macular degeneration (nAMD).

    Methods: 28 eyes of 26 patients with nAMD were imaged using conventional time domain (TD) OCT as well as three spectral-domain (SD) OCT systems. Radial scans of 6mm in size were compared between Stratus and Topcon OCT, in addition to raster scans of all three SD-OCT devices. Retinal thickness values were analyzed.

    Results: Spectralis SD-OCT demonstrated highest values of all OCT devices in central millimetre thickness (CMMT) and Topcon OCT raster scans showed lowest values. Significant correlations could be found between the central mm thickness (CMMT) measurements of Cirrus and Spectralis OCT (r=0.87). Analyses showed best segmentation for Cirrus and Spectralis SD-OCTs. Cirrus 200x200x1024 scans showed 4% and Stratus OCT 38% moderate or severe segmentation errors.

    Conclusion: This study showed that SD-OCT presents retinal morphology in greatest detail and is more robust to the typical segmentation artefacts than TD-OCT. Retinal thickness values were generally higher in SD-OCT analysis. Different performances of automatic retinal thickness analysis indicate the potential of different software algorithms to quantify retinal morphology in nAMD. Further development of current algorithms may improve quantification of retinal thickness detection in the future even further.

    Statistics from

    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.

    Linked Articles

    • At a glance
      Harminder S Dua Arun D Singh