Automated layer segmentation of optical coherence tomography images

OCT images have now become a very popular topic in the field of image processing. By measuring the retinal nerve fiber layer thickness, such diseases like glaucoma or cataract can be diagnosed. An automated boundary segmentation algorithm is proposed for fast and reliable quantification of six intra-retinal boundaries in optical coherence tomography (OCT) images. The algorithm includes four steps. First of all, the image will be filtered by a bilateral filter which suppress the local image noise but keep the global image variation across the retinal layer boundary. Secondly, the image will be cut into several nonvessel sections according to the retinal blood vessels. Thirdly, segmentations in different layers based on gradient information are utilized. In the end, a shortest path search is applied to optimize the edge selection. The segmentation algorithm was validated with independent manual segmentation and demonstrates high accuracy and reproducibility in segmenting normal OCT volumes.