International Journal of Advances in Computer Science and Its Applications
Author(s) : M.DHANALAKSHMI , R.C.THARANI, R.SHANTHINI, S.PONMALAR
The Optical Coherence Tomography (OCT) is an emerging technology in which lot of research work is going on to simplify the diagnosis of eye diseases. The proposed work is aimed at developing a computer aided diagnostic tool for detection of macular disorders like Age Related Macular Degeneration (ARMD) and its severity from which the dosage level and frequency of the intravitreal Anti-Vascular endothelial growth factor agents like Ranibizumab and Bevacizumab given to treat ARMD. The Retinal 3D OCT images are preprocessed using various filters and based on the parameters (MSE, RMSE, PSNR,UIQI ) observed, Shock Filter gave the most prominent result The Symptomatic Exudate-Associated Derangement(SEAD) region is then segmented by Active Contour method. The features extracted are Runlength features, intensity features and Co-occurrence matrix features. Extreme Learning Machine (ELM), Self-Adaptive Resource Allocation Network (SRAN) and Meta-Cognitive Neural Network (McNN) are trained to classify the severity of the disease based on these extracted features and overall efficiency of 80, 87 and 80 percent is achieved respectively.