Journal
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
Volume 46, Issue -, Pages 315-327Publisher
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.compmedimag.2015.09.008
Keywords
Optical coherence tomography (OCT); Choroidal thickness; Choroidal volume; Structural similarity (SSIM) index; Tensor voting
Funding
- Department of Electronics and Information Technology (Deity), Govt. of India [13(6)/2010-CCBT]
Ask authors/readers for more resources
A variety of vision ailments are indicated by anomalies in the choroid layer of the posterior visual section. Consequently, choroidal thickness and volume measurements, usually performed by experts based on optical coherence tomography (OCT) images, have assumed diagnostic significance. Now, to save precious expert time, it has become imperative to develop automated methods. To this end, one requires choroid outer boundary (COB) detection as a crucial step, where difficulty arises as the COB divides the choroidal granularity and the scleral uniformity only notionally, without marked brightness variation. In this backdrop, we measure the structural dissimilarity between choroid and sclera by structural similarity (SSIM) index, and hence estimate the COB by thresholding. Subsequently, smooth COB estimates, mimicking manual delineation, are obtained using tensor voting. On five datasets, each consisting of 97 adult OCT Bscans, automated and manual segmentation results agree visually. We also demonstrate close statistical match (greater than 99.6% correlation) between choroidal thickness distributions obtained algorithmically and manually. Further, quantitative superiority of our method is established over existing results by respective factors of 27.67% and 76.04% in two quotient measures defined relative to observer repeatability. Finally, automated choroidal volume estimation, being attempted for the first time, also yields results in close agreement with that of manual methods. (c) 2015 Elsevier Ltd. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available