Display Settings:

Format

Send to:

Choose Destination
See comment in PubMed Commons below
Biomed Opt Express. 2012 Jan 1;3(1):86-103. doi: 10.1364/BOE.3.000086. Epub 2011 Dec 12.

Automated choroidal segmentation of 1060 nm OCT in healthy and pathologic eyes using a statistical model.

Author information

  • 1Center for Medical Physics and Biomedical Engineering, Medical University Vienna, General Hospital Vienna 4L, Waehringer Guertel 18-20, A-1090 Vienna, Austria. kajicv1@cf.ac.uk

Abstract

A two stage statistical model based on texture and shape for fully automatic choroidal segmentation of normal and pathologic eyes obtained by a 1060 nm optical coherence tomography (OCT) system is developed. A novel dynamic programming approach is implemented to determine location of the retinal pigment epithelium/ Bruch's membrane /choriocapillaris (RBC) boundary. The choroid-sclera interface (CSI) is segmented using a statistical model. The algorithm is robust even in presence of speckle noise, low signal (thick choroid), retinal pigment epithelium (RPE) detachments and atrophy, drusen, shadowing and other artifacts. Evaluation against a set of 871 manually segmented cross-sectional scans from 12 eyes achieves an average error rate of 13%, computed per tomogram as a ratio of incorrectly classified pixels and the total layer surface. For the first time a fully automatic choroidal segmentation algorithm is successfully applied to a wide range of clinical volumetric OCT data.

2011 Optical Society of America

KEYWORDS:

(100.0100) Image processing; (100.3008) Image recognition, algorithms and filters; (170.4500) Optical coherence tomography; (170.4580) Optical diagnostics for medicine

PMID:
22254171
[PubMed]
PMCID:
PMC3255345
Free PMC Article

Images from this publication.See all images (11)Free text

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for PubMed Central
    Loading ...
    Write to the Help Desk