Article Text

Download PDFPDF
Letters
Standardised disturbance of the optical coherence tomography signal has varying effects on the scan quality assessment when comparing four devices
  1. Pauline H B Kok1,
  2. Hille W van Dijk1,
  3. Marilette Stehouwer1,
  4. Thomas J T P van den Berg2,
  5. R O Schlingemann1,
  6. Frank D Verbraak1,3
  1. 1 Department of Ophthalmology, Academic Medical Center, Amsterdam, The Netherlands
  2. 2 Netherlands Institute for Neuroscience, Royal Academy of the Netherlands, Amsterdam, The Netherlands
  3. 3 Biomedical Engineering and Physics, Academic Medical Center, Amsterdam, The Netherlands
  1. Correspondence to Pauline H B Kok, Department of Ophthalmology, Room: D2-420, Academic Medical Center, University of Amsterdam, Meibergdreef 9, Amsterdam 1105 AZ, The Netherlands; p.h.kok{at}amc.nl

Statistics from Altmetric.com

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.

Introduction

Optical coherence tomography (OCT) has become a successful application in ophthalmology. Time-domain OCT (Stratus, Carl Zeiss Meditec) has been overtaken by spectral-domain OCT (SDOCT) which has major advances in imaging speed, sensitivity and image resolution.1 ,2 Several SDOCT devices are commercially available.

In this small study, we assessed the change in subjective image quality and image quality parameter (IQP) provided by four different SDOCT devices using artificial filters simulating optical eye media disturbances.

Methods

In four healthy subjects, single non-averaged B-scans of the macula were acquired using four commercially available SDOCT systems:

  1. 3D OCT-1000 MarkII (Topcon Medical Systems, Inc, Oakland, New Jersey, USA): Software V.3.21, 27.000 A-scans per second, superluminescent diode 840 nm light source, 5–6 micron axial resolution, ‘Q-factor’ IQP scale 0–100.

  2. Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, California, USA): Software V.2.0, 27.000 A-scans per second, superluminescent diode 840 nm light source, 5 micron axial resolution, ‘Signal Strength’ IQP scale 0–10.

  3. RTVue OCT (Optovue, Inc, Fremont, California, USA): Software V.2.0.4.0, 26.000 A-scans per second, superluminescent diode 840 nm light source, 5 micron axial resolution, ‘Signal Strength Index’ IQP scale 0–100.

  4. Spectralis OCT (Heidelberg Engineering, GmbH, Heidelberg, Germany): Software V.5.1.3, 40.000 A-scans per second, superluminescent diode 820 nm light source, 7 micron axial resolution, ‘Quality’ IQP scale 0–10.

Four series of artificial filters, with a reflective, an absorbing (both attenuating), a scattering (straylight), and refractive (defocusing) property, respectively, were placed in front of the eye while scanning 20 filters in total. The effective optical density (OD) of the filters, ranging from low to high, was calculated as described previously3 and compared with the IQP of the OCT images as supplied by the different devices.

Results

For each SDOCT device, a rather precise linear relationship was found between the optical density of the filters and the change in the IQP of the device (Q-factor, signal strength, signal strength index and quality), but with different slopes (see figure 1, p<0.0001). This indicates that light attenuation is the main factor determining the IQP provided, but also that the IQPs are not directly comparable. Moreover, figure 1 shows that differences between filter types are small.

Figure 1

Correlation between the percentage change in image quality parameter and the optical density of the filters used. Pearson correlation 3D optical coherence tomography (OCT)-1000 MarkII R2=0.98, p<0.0001; Cirrus R2=0.93, p<0.0001; RTVue R2=0.93, p<0.0001; Spectralis, R2=0.82, p<0.0001.

Figure 2 shows OCT images made with the four devices, without filter, through an identical filter and through a filter selected to result in an image with identical, barely distinguishable intraretinal structures. Images made through the same filter (OD=0.54) demonstrated differences in visualisation of the preretinal, intraretinal and subretinal structures, and background noise. OD values of the selected images with identical, barely distinguishable intraretinal structures varied between instruments from OD 0.54 to OD 0.75.

Figure 2

Optical coherence tomography (OCT) images made with the four devices, without filter, through an identical filter (optical density (OD)=0.54) and through a filter resulting in an image where intraretinal structures can be just distinguished.

Discussion

As expected, there was a rather precise linear relationship between the optical density of the filters and the IQPs, confirming that attenuation of the OCT signal is the main determinant of the IQP calculation. The observed differences in the effect of a series of filters with different OD on the subjective quality of images and on the IQPs are probably caused by differences in hardware, acquisition methods and postacquisition processing of the signal. The range of quality measures and the basis for image quality assessment are not comparable across instruments.4 Note that we used single non-averaged B-scans, whereas averaging processes and eye tracking improves image quality, which is for example used in standard Spectralis OCT volume scanning.

In conclusion, equal disturbances of the OCT signal have different effects on subjective scan quality and the IQP provided by the instrument in the four SDOCT devices investigated. Standardisation of the IQP, which would be desirable in order to compare different devices and study results, seems to be impossible.

Acknowledgments

The authors thank the Rotterdams Oogheelkundig Instituut (R.O.I.) and the Department of Ophthalmology, Flevoziekenhuis Almere, for the collaboration.

References

Footnotes

  • Contributors PK: first draft, substantial contribution to conception and design, acquisition of data and analysis and interpretation of the data. HD and MS: substantial contributions to acquisition of data. TB: substantial contributions to conception and design and revising the article critically for important intellectual content; technical supervisor as physicist. RS: revising the article critically for important intellectual content; supervisor as medical retina specialist. FD: substantial contributions to conception and design and final approval of the version to be published.

  • Competing interests None.

  • Ethics approval Ethics committee of the Academic Medical Center in Amsterdam.

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