August 2019
Volume 60, Issue 11
Free
ARVO Imaging in the Eye Conference Abstract  |   August 2019
A universal three-dimensional registration algorithm on OCT/OCTA for speckle reduction and visualization
Author Affiliations & Notes
  • Yuxuan Cheng
    Bioengineering, University of Washington, Seattle, Washington, United States
  • Qinqin Zhang
    Bioengineering, University of Washington, Seattle, Washington, United States
  • Shaozhen Song
    Bioengineering, University of Washington, Seattle, Washington, United States
  • Zhongdi Chu
    Bioengineering, University of Washington, Seattle, Washington, United States
  • Ruikang Wang
    Bioengineering, University of Washington, Seattle, Washington, United States
  • Footnotes
    Commercial Relationships   Yuxuan Cheng, None; Qinqin Zhang, None; Shaozhen Song, None; Zhongdi Chu, None; Ruikang Wang, None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science August 2019, Vol.60, 025. doi:https://doi.org/
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Yuxuan Cheng, Qinqin Zhang, Shaozhen Song, Zhongdi Chu, Ruikang Wang; A universal three-dimensional registration algorithm on OCT/OCTA for speckle reduction and visualization. Invest. Ophthalmol. Vis. Sci. 2019;60(11):025. doi: https://doi.org/.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : OCT images often suffer from speckle noises and motion-induced artifacts that affect its accurate, reproducible, and reliable assessment of the samples To mitigate this challenge, we propose a method based on a fast, robust and three-dimensional registration method to enhance the contrast of both OCT and OCT-Angiography(OCT-A) images without pre-processing.

Methods : All data were collected by a commercial SSOCTA device (PLEX® Elite 9000 (ZEISS, Dublin, CA)). 6x6 mm2 scans were utilized to test our method with a dimension 500x500x1560 (x, y, z). 10 repeated volumes were acquired on all participants. The 3D affine transformation was first implemented to low resolution images to correct the bulk motion between different OCT structural volumes and then use the non-ridge B-spline registration to refine the local alignment between the reference and moving volumes on both OCT structural and flow images. Finally, the corrected images were weight-averaged with the previous results and are assigned as the reference images for the next registration. The registration program was developed based on Matlab platform and it costs about 30 minutes for 10 volumes averaging on workstation configured with the Intel Xeon 2630-v3 CPU and 128G RAM.

Results : Five subjects have been enrolled in this study with 10 repeated volume scans either with 3x3 mm2 scan or 6x6 mm2 scan. The deformation of sequential images can be rapidly corrected and the motion artifacts can be effectively suppressed. The average peak signal-to-noise ratio (PSNR) measures the misalignment of images, which increased from 25 dB to 72dB.

Conclusions : The averaged results from both B-scan and the en-face projection of retinal layer and choriocapillaris showed an enhanced SNR and vascular connectivity comparing to the single scan image. This approach provides better visualization of microstructures and microvasculature that could be helpful in monitoring the disease progression and further assistant the therapeutic intervention.

This abstract was presented at the 2019 ARVO Imaging in the Eye Conference, held in Vancouver, Canada, April 26-27, 2019.

 

Figure 1.OCTA images of original single and averaged images. (a-c) retinal layer; (d-f) choriocapillaris layer; (g-i) the cross-sectional image located at the yellow dashed line.
Figure 2 (a) The relationship between SNR of OCT structure images and the number of averaging. The dashed line represents the fitting curve by the square of the number of repeated volumes.

Figure 1.OCTA images of original single and averaged images. (a-c) retinal layer; (d-f) choriocapillaris layer; (g-i) the cross-sectional image located at the yellow dashed line.
Figure 2 (a) The relationship between SNR of OCT structure images and the number of averaging. The dashed line represents the fitting curve by the square of the number of repeated volumes.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×