ILM and Fovea Detection using Standard Deviation Profiling Method

Abstract

Fovea located in the Macular region of the retina is the important location of the eye that is responsible for vision. Fovea can be observed from optical coherence tomography (OCT) images. This type of medical image is actively being used in the medical field to detect ocular diseases such as Age-related Macular Degeneration and Diabetic Retinopathy. However, it is often challenging to spot the fovea in abnormal OCT images for diagnostic purposes. In this paper, we proposed a method called standard deviation profiling to detect the Inner Limiting Membrane (ILM). Features extracted from the ILM layer were used in the decision tree for case classification. The fovea was detected from the ILM layer based on a rule-based method. For the ILM detection, the results show that it can significantly reduce the root mean square error compared with the CASEREL and Canny edge detection methods. For fovea detection, we achieve an overall accuracy of 94%.