Systems and methods to compensate for reflectance variation in oct angiography
AssigneeOregon Health & Science University
InventorsDavid Huang, Yali Jia, Simon Gao
Life of this patent
Estimating the term…
Abstract
Methods and systems for improving quantification of OCT angiography data are disclosed. The disclosure specifically relates to methods for compensating for the effect of tissue reflectance to improve the accuracy and repeatability of OCT angiography measurements. These improvements are effected by deriving and then utilizing a dynamic thresholding approach to process decorrelation data to properly classify flow versus non-flow data in OCT angiograms. The disclosed methods overcome quantification errors associated with within-scan variations in reflectance as well as repeatability problems associated with differences in scan quality over successive imaging sessions.
Claims21
A method of generating a dynamic thresholding equation for use in optical coherence tomography (OCT) angiography, the method comprising:
- for individual participants of a set of participants:
- acquiring a structural OCT dataset from an avascular region of the retina, the structural OCT dataset comprising reflectance values;
- calculating an OCT decorrelation dataset from the avascular region of the retina, the OCT decorrelation dataset comprising decorrelation values;
- fitting a first function to mean decorrelation values versus mean reflectance values obtained from the set of participants;
- fitting a second function to a standard deviation of decorrelation values versus mean reflectance values obtained from the set of participants;
- and
- generating a dynamic thresholding equation from the first function and second function, the dynamic thresholding equation to classify pixels of the OCT decorrelation dataset as flow pixels or no flow pixels.
A method of compensating for the effect of tissue reflectance in optical coherence tomography (OCT) angiography, the method comprising:
- receiving a structural OCT dataset;
- obtaining an OCT decorrelation dataset;
- generating an en face angiogram from the OCT decorrelation dataset, the en face angiogram including a plurality of pixels;
- generating an en face reflectance image from the structural OCT dataset, the en face reflectance image including a plurality of pixels;
- for individual pixels in the en face reflectance image:
- calculating a dynamic threshold value based on the pixel's reflectance value;
- classifying a corresponding pixel in the en face angiogram as “flow,” provided that the pixel's decorrelation value is greater than the dynamic threshold value;
- classifying the corresponding pixel in the en face angiogram as “no flow,” provided that the pixel's decorrelation value is less than the dynamic threshold value;
- and
- returning a reflectance-compensated decorrelation image.
Dependencies are inferred from the claim text.