Automated Retinal Layer Segmentation
OCT allows a direct visualization of the retinal background and its layered structure. This proved to be of substantial benefit for the glaucoma disease research. During glaucoma progression the supporting tissue and nerve fibers in the retina are lost. Thus a thinning of the retinas innermost layer, the retinal nerve fiber layer (RNFL) is observed. Since the first appearance of commercially available OCT systems automated retinal layer segmentation algorithms were presented to objectively quantify this thickness loss. We develop a retinal layer segmentation algorithm with the following goals:
- Robust segmentation of the RNFL. The results have to be reproducible and precise.
- No patient dependent parameter change. The same parameter set should be used for normal subjects and glaucomatous patients. The algorithm should cope with RNFL thinning and also loss.
- Segmentation of 2D circular and linear scans as well as volume scans are possible.
- Exploit the results for better visualizations of the images (e.g., geometry corrections, RNFL maps)
Original circular OCT B-Scan image of a glaucoma patient with local RNFL loss.<br> |
Result of our nerve fiber layer segmentation. |
Currently published journal articles:
- Markus A. Mayer, Ralf. P. Tornow, Christian Y Mardin, Ralf P. Tornow: "Retinal Nerve Fiber Layer Segmentation on FD-OCT Scans of Normal Subjects and Glaucoma Patients", Biomedical Optics Express 1 (2010) No. 5 pp. 1358-1383. Article Download
Early versions of a segmentation algorithm for circular OCT scans based on fuzzy C-means clustering were published:
- Markus A. Mayer, Ralf. P. Tornow, RĂ¼diger Bock, Joachim Hornegger, Friedrich E. Kruse: "Automatic Nerve Fiber Layer Segmentation and Geometry Correction And Geometry Correction On Spectral Domain OCT Images Using Fuzzy C-means Clustering", The Association for Research in Vision and Ophthalmology, Inc. (ARVO) (Annual Meeting) in Fort Lauderdale, Florida, USA, 2008. Poster Download
- Markus A. Mayer, Ralf P. Tornow, Joachim Hornegger, Friedrich E. Kruse:
"Fuzzy C-means Clustering For Retinal Layer Segmentation On High Resolution OCT Images", Proceedings of the Biosignal 2008 International Eurasip Conference (Biosignal 2008), Brno, Czech Republic, Vol. 19, 2008. Paper Download
For this project, the OCTSEG Software for segmenting and visualizing OCT data was written. This software is downloadable and may be used for free.