April 2014
Volume 55, Issue 13
Free
ARVO Annual Meeting Abstract  |   April 2014
Afocal AO-OCT Image Quality Improvements through 3D Registration and Averaging
Author Affiliations & Notes
  • Ravi S Jonnal
    Ophthalmology, UC Davis, Sacramento, CA
    Vision Science and Advanced Retinal Imaging Laboratory (VSRI), UC Davis, Sacramento, CA
  • Justin V Migacz
    Ophthalmology, UC Davis, Sacramento, CA
    Vision Science and Advanced Retinal Imaging Laboratory (VSRI), UC Davis, Sacramento, CA
  • Robert J Zawadzki
    Ophthalmology, UC Davis, Sacramento, CA
    Vision Science and Advanced Retinal Imaging Laboratory (VSRI), UC Davis, Sacramento, CA
  • Sang Hyuck Lee
    Ophthalmology, UC Davis, Sacramento, CA
    Vision Science and Advanced Retinal Imaging Laboratory (VSRI), UC Davis, Sacramento, CA
  • John S Werner
    Ophthalmology, UC Davis, Sacramento, CA
    Vision Science and Advanced Retinal Imaging Laboratory (VSRI), UC Davis, Sacramento, CA
  • Footnotes
    Commercial Relationships Ravi Jonnal, US 7,364,296 (P); Justin Migacz, None; Robert Zawadzki, None; Sang Hyuck Lee, None; John Werner, None
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science April 2014, Vol.55, 5204. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Ravi S Jonnal, Justin V Migacz, Robert J Zawadzki, Sang Hyuck Lee, John S Werner; Afocal AO-OCT Image Quality Improvements through 3D Registration and Averaging. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5204.

      Download citation file:


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

      ×
  • Supplements
Abstract
 
Purpose
 

Adaptive optics (AO) has permitted the highest resolution images of the living human retina to date. We developed a spherical-mirror-based AO-OCT system designed to compensate for off-axis aberrations and pupil wander by bringing the beam out of a single plane. The system was used for fast acquisition of series of volumetric images of the retina. Here we show that these volumes may be registered and averaged to improve image quality. The precision of such registration and averaging is limited ultimately by blur and speckle, an unavoidable source of noise in coherent imaging. This design is expected to improve resolution and reduce speckle size, thereby fostering better registration and averaging, for addressing scientific hypotheses and clinical application.

 
Methods
 

The afocal spectral domain (SD) AO-OCT system has been described previously. The AO system ran continuously (25 fps) in closed loop during the course of imaging. Axial and lateral resolutions are ~3.0 and ~2.5 μm, respectively. Two subjects without known ocular disease were imaged at several retinal locations. At each, a series of five volumes was acquired within 5 s. Volumes were segmented using a model-based approach. This permitted extraction of areal projections of layers of interest. Next, considering the granularity and contrast of these projections, as well as eye movement artifacts present in the images, a suitable reference image was selected--typically a projection of the photoreceptor mosaic. A strip-based registration technique was applied to register corresponding projections from the other volumes to the reference image. The output of the registration algorithm was then applied to the whole volumes.

 
Results
 

Qualitative improvement in the appearance of averaged images (right) over single images (left) of the nerve fiber layer (top) and inner plexiform layer (bottom) is shown for one subject (7 deg nasal to the fovea). Images subtended ~250 μm in each dimension. The improvements in image quality resulted, likely, from reductions in speckle and photon noise due to averaging and associated improvement in SNR.

 
Conclusions
 

With the development of an afocal AO-OCT system, we have increased the resolution of the system by improving the AO system's sensitivity to wavefront error. Here we demonstrate preliminary results of a technique to mitigate some of the inherent constraints on signal, namely speckle and photon noise, by averaging volumes.

  
Keywords: 552 imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • 549 image processing  
×
×

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.

×