PROJECT TITLE :

High Accuracy Retinal Layer Segmentation For Optical Coherence Tomography Using Tracking Kernels Based On Gaussian Mixture Model - 2014

ABSTRACT:

Ophthalmology needs automated segmentation of retinal layers in optical coherence tomography images to supply valuable disease information. Sensitive extraction of correct layer boundaries stable against local image quality degradation is necessary. We propose and demonstrate a strong, accurate segmentation methodology with high stability and sensitivity. The method uses an intelligent tracking kernel and a clustering mask based mostly on the Gaussian mixture model (GMM). Combining these ideas yields robust, degradation-free tracking with highly sensitive pixel classification. The kernel extracts boundaries by moving and matching its double faces with regionally clustered images generated by GMM clustering. The cluster-guided motion of the kernel permits sensitive classification of structures on a single-pixel scale. This system targets seven major retinal boundaries. Then, using peak detection, additional 2 straightforward boundaries are easily grabbed in regions where their distinct features emerge sufficiently within the limited house remaining when the previous segmentation. Using these hybrid modes, successful segmentation of 9 boundaries of eight retinal layers in foveal areas is demonstrated. A zero.909 fraction of a pixel difference seems between boundaries segmented manually and using our algorithm. Our technique was developed for use with low-quality information, permitting its application in numerous morphological segmentation technologies.