September 2016
Volume 57, Issue 12
Open Access
ARVO Annual Meeting Abstract  |   September 2016
A novel approach to reducing decorrelation tail artifacts in OCT angiography
Author Affiliations & Notes
  • Homayoun Bagherinia
    Carl Zeiss Meditec, Inc., Dublin, California, United States
  • Jesse J Jung
    East Bay Retina Consultants, Inc., Oakland, California, United States
  • Soraya Rofagha
    East Bay Retina Consultants, Inc., Oakland, California, United States
  • Patty Chung
    East Bay Retina Consultants, Inc., Oakland, California, United States
  • Scott Lee
    East Bay Retina Consultants, Inc., Oakland, California, United States
  • Michael Chen
    Carl Zeiss Meditec, Inc., Dublin, California, United States
  • Footnotes
    Commercial Relationships   Homayoun Bagherinia, Carl Zeiss Meditec, Inc (E); Jesse Jung, Carl Zeiss Meditec, Inc (C); Soraya Rofagha, Carl Zeiss Meditec, Inc (C); Patty Chung, Carl Zeiss Meditec, Inc (C); Scott Lee, Carl Zeiss Meditec, Inc (C); Michael Chen, Carl Zeiss Meditec, Inc (C)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science September 2016, Vol.57, 449. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Homayoun Bagherinia, Jesse J Jung, Soraya Rofagha, Patty Chung, Scott Lee, Michael Chen; A novel approach to reducing decorrelation tail artifacts in OCT angiography. Invest. Ophthalmol. Vis. Sci. 2016;57(12):449.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : To describe a mathematical approach to reducing decorrelation artifacts in OCT angiography (OCTA) images.

Methods : OCTA is prone to decorrelation tail artifacts due to the high scattering property of blood within overlying patent vessels, creating artifacts that interfere with the interpretation of OCTA images. We present a new approach for reducing the decorrelation tail artifacts based on an inverse problem estimation using two slab images generated from two or more retinal layers. A slab image of p=m×n can be generated by integrating, or other techniques to select a representative value of the cube motion contrast data within one or more layers. In the inverse problem approach discussed here, it is assumed that a lower slab image is generated as a result of mixture of an upper slab (a sub-volume above the lower layer) and the lower slab image without decorrelation tail artifacts (the unknown image to reconstruct). Assuming a multiplicative mixing model, an inverse formulation of decorrelation tail problem is minx ∥WAx-b∥2+α∥Γx∥2, where A is a diagonal p×p matrix whose diagonal elements are the pixel values of the upper slab image, W is a diagonal p×p matrix containing the weight values for each pixel of the upper slab image, b is a column vector of p whose elements are the pixel values of the lower slab image, α is the regularization parameter, Γ is a matrix of p×p which contains filters that enforce smoothness. x is a column vector of p whose elements are the pixel values of the lower slab image with reduced decorrelation tail artifacts.

Results : The figures below show examples of decorrelation artifacts reduction by our method for macular telangiectasia (Mac Tel) type 2, and subretinal neovascular membrane (SRNVM) cases before and after reducing the decorrelation artifacts using CIRRUS (ZEISS Dublin, CA).

Conclusions : We have presented a mathematical approach to reduce the inner and outer retinal decorrelation tail artifacts in OCTA images. Our approach can be utilized for visualization and interpretation of various disease cases such as SRNVM and Mac Tel.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.

 

Mac Tel type 2 examples from left to right: superficial capillary plexus, deep capillary plexus (DCP) with corresponding B-scans; and DCP with reduced decorrelation artifacts.

Mac Tel type 2 examples from left to right: superficial capillary plexus, deep capillary plexus (DCP) with corresponding B-scans; and DCP with reduced decorrelation artifacts.

 

SRNVM case from left to right: DCP and custom slab segmenting the type 2 neovascular membrane with corresponding B-scabs; and the custom slab with reduced decorrelation artifacts.

SRNVM case from left to right: DCP and custom slab segmenting the type 2 neovascular membrane with corresponding B-scabs; and the custom slab with reduced decorrelation artifacts.

×
×

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.

×