Automated segmentation of retinal layers from OCT images using structure tensor and kernel regression+GTDP approach
Autor: | Samra Naz, M. Usman Akram, Shoab A. Khan |
---|---|
Rok vydání: | 2017 |
Předmět: |
Retina
genetic structures medicine.diagnostic_test Computer science business.industry Retinal Graph theory Structure tensor eye diseases Edge detection chemistry.chemical_compound medicine.anatomical_structure Optical coherence tomography chemistry medicine Kernel regression Computer vision Segmentation sense organs Artificial intelligence business |
Zdroj: | 2017 1st International Conference on Next Generation Computing Applications (NextComp). |
DOI: | 10.1109/nextcomp.2017.8016182 |
Popis: | Macular edema (ME) is trending to be an increasing disease among the working adult group that causes retinal swelling within sub-retinal layers due to accumulation of protein deposits. Optical coherence tomography (OCT) imaging technique gives a complete cross-sectional view of the retina and helps in early detection of ME. In this paper, we explore different techniques for accurate and reliable segmentation of retinal layers. It uses advance edge detection technique of structure tensor for accurate layer segmentation. Structure tensor algorithm is compared with kernel regression and GTDP (Graph Theory and Dynamic Programming) based layer segmentation technique. The evaluation is performed using a subset of publicly available OCT image dataset of 10 patients containing 108 OCT scans which is provided by Duke University. |
Databáze: | OpenAIRE |
Externí odkaz: |