April 2011
Volume 52, Issue 14
ARVO Annual Meeting Abstract  |   April 2011
High-resolution Adaptive Optics Retinal Imaging With Slope Extrapolation Boundary Treatment
Author Affiliations & Notes
  • Weiyao Zou
    School of Optometry, Indiana University, Bloomington, Indiana
  • Xiaofeng Qi
    School of Optometry, Indiana University, Bloomington, Indiana
  • Gang Huang
    School of Optometry, Indiana University, Bloomington, Indiana
  • Stephen A. Burns
    School of Optometry, Indiana University, Bloomington, Indiana
  • Footnotes
    Commercial Relationships  Weiyao Zou, None; Xiaofeng Qi, None; Gang Huang, None; Stephen A. Burns, None
  • Footnotes
    Support  NIH Grants EY04395, EY14375, P30EY019008.
Investigative Ophthalmology & Visual Science April 2011, Vol.52, 4060. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Weiyao Zou, Xiaofeng Qi, Gang Huang, Stephen A. Burns; High-resolution Adaptive Optics Retinal Imaging With Slope Extrapolation Boundary Treatment. Invest. Ophthalmol. Vis. Sci. 2011;52(14):4060.

      Download citation file:

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

  • Supplements

Most algorithms for Adaptive Optics (AO) retinal imaging are limited in aberration correction due to edge effects. We implemented a technique to improve AO control accuracy by improving the slope boundary condition with a dual deformable-mirror (DM) AO scanning laser ophthalmoscope.


The problem of wavefront control is to estimate the wavefront phase from slope measurements and then control the DMs to correct the phase. Because wavefront slope measurements usually have large errors near pupil edges and a slope variation at pupil edge can dramatically change the wavefront estimation result (Roddier 1993), so AO control is very sensitive to boundary slope errors. In addition, human pupils tend to change during retinal imaging. Thus, handling the boundary conditions is a key step in providing robust human wavefront control. To reduce the edge effect we added two steps to our AO control. First, for avoiding large boundary slope errors we calibrated a series of influence matrices for different pupil sizes, so that we can use a large influence matrix for a large pupil and a small influence matrix for a small pupil. Second, for handling pupil dynamics we designed an algorithm to control a larger size pupil relative to the subject’s pupil. A Zernike modal fit was performed to the slope data within the pupil and then extrapolated across the pupil boundary. Thus, the presence of slope estimates outside of the real data helps stabilizing the pupil. As the subject pupil moves within the control pupil, a stable high-resolution AO imaging should be guaranteed.


Fig.1 shows the AOLSO imaging performance with/without the boundary slope extrapolation for an artificial eye. For the control pupil sizes smaller than the physical pupil size, the wavefront control accuracy is unaffected. When the control pupil size is larger, the AO control without slope extrapolation deteriorates, but the new algorithm avoids this degradation beyond the physical pupil up to 1-mm larger.


The slope-extrapolation approach was tested for stabilizing high accuracy AO performance for moving and fluctuating pupils, which is important in clinical retinal imaging.  

Keywords: imaging/image analysis: non-clinical • imaging/image analysis: clinical • aberrations 

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.