June 2022
Volume 63, Issue 7
Open Access
ARVO Annual Meeting Abstract  |   June 2022
Automated correction of jitter, interleaving and non-uniform image sampling in scanning ophthalmoscopes
Author Affiliations & Notes
  • Alfredo Dubra
    Ophthalmology, Stanford University, Stanford, California, United States
  • Vyas Akondi
    Ophthalmology, Stanford University, Stanford, California, United States
  • Aubrey Hargrave
    Ophthalmology, Stanford University, Stanford, California, United States
  • Bartlomiej Kowalski
    Ophthalmology, Stanford University, Stanford, California, United States
  • Footnotes
    Commercial Relationships   Alfredo Dubra None; Vyas Akondi None; Aubrey Hargrave None; Bartlomiej Kowalski None
  • Footnotes
    Support  Research to Prevent Blindness (Departmental award); National Eye Institute (P30EY026877, R01EY025231, R01EY031360, R01EY032147, R01EY032669).
Investigative Ophthalmology & Visual Science June 2022, Vol.63, 4457 – F0136. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Alfredo Dubra, Vyas Akondi, Aubrey Hargrave, Bartlomiej Kowalski; Automated correction of jitter, interleaving and non-uniform image sampling in scanning ophthalmoscopes. Invest. Ophthalmol. Vis. Sci. 2022;63(7):4457 – F0136.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Sampling jitter and non-uniform scanning angular velocity limit the accuracy and precision of scanning ophthalmoscopes. Here we explore the mitigation of these problems through the recording of the optical scanners’ orientation in synchrony with the light detection, followed by resampling of retinal image.

Methods : Distortion grid and photoreceptor mosaic images were captured using a custom scanning light ophthalmoscope modified to record scanner orientation for each image pixel. The horizontal (resonant) scanner orientation samples were fitted to a cosine model for each image line to correct image warping and sampling jitter through image data resampling, as well as to interleave images collected during the clockwise and counterclockwise portions of the scanner rotation. This was followed by image warp correction due to non-uniform angular speed of the vertical scanner through a second image resampling. Jitter correction was evaluated using normalized cross-correlation within each line of images from a Ronchi ruling, and as the norm of standard deviation maps from a stack of co-registered retinal images.

Results : The fitting of the cosine models showed ~0.36 pix jitter (STD) which reduced jitter by more than a factor of 6 after resampling and allowed interleaving image lines captured during the counter-rotating portions of the horizontal scan cycle with sub-pixel precision. The norm of the standard deviation maps, a registration error metric, decreased by ~10-15% with jitter correction and ~16-20% with jitter and vertical warping correction in the distortion grid images. More modest ~1 and ~7% respective reductions were observed in a small set of photoreceptor images affected by eye rotation (not corrected by the registration).

Conclusions : The proposed jitter and warp correction in retinal images from scanning ophthalmoscopes can improve the sensitivity of retinal imaging biomarkers by producing more anatomically truthful images. In adaptive optics scanning ophthalmoscopes, these corrections can also help to reduce the number of images that is captured at each retinal location for subsequent averaging, necessary to increase signal to noise ratio (SNR), and a most important barrier to the translation of this technology to the clinic. This approach could also increase SNR in angiograms from optical coherence tomography.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

×
×

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.

×