الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 158 |  |  |
| | | from | 158 | | |
Abstract
First, an automated comprehensive computer-aided diagnostic (CAD) system is proposed for AMD diagnosis. The proposed CAD system is capable of deriving clinically meaningful features from optical coherence tomography (OCT) B-scan images to differentiate between normal retina, different grades of AMD (early, intermediate, geographic atrophy (GA), inactive wet or active neovascular disease (exudative or wet AMD)), and non-AMD diseases. Particularly, the system extract retinal OCT-based imaging markers that are correlated with the progression of AMD, which include: (i ) subretinal tissue, sub-retinal pigment epithelial tissue, ntraretinal fluid, subretinal fluid, and choroidal hypertransmission detection using the DeepLabV3+ with backbone ResNet50 network; (ii ) detection of merged retina layers by using a novel convolutional neural network model; (iii ) drusen detection based on the analysis of the estimated 2D curvature; (iv ) estimation of retinal layers’ thickness, and first-order and second-order reflectivity features derived from a Markov-Gibbs random field model and the gray level co-occurrence matrix. Those features are used to grade a retinal OCT in a hierarchical decision tree process. The first step looks for severe disruption of retinal layers’ indicative of advanced AMD. These cases are analyzed further to diagnose GA, inactive wet AMD, active wet AMD, and non-AMD diseases. Less severe cases are analyzed using a different pipeline to identify OCT with AMD-specific pathology, which are graded as intermediate-stage or early-stage AMD. The remainder is classified as either being a normal retina or having other non-AMD pathology. The proposed system in the multi-way classification task, evaluated on 1285 OCT images from two different tertiary referral centers in the U.S., the University of Louisville (UofL) and Legacy Devers Eye Institute (LDEI), achieved 90.82% accuracy. we evaluated our system based on each class outcomes against all other classes. Thus, we used three evaluation metrics, i.e., Precision, Recall, F1-score for each individual class. These promising results demonstrated the capability to automatically distinguish between normal eyes and all AMD grades in addition to non-AMD diseases. Second, an OCT-CAD method is proposed to detect DR early using structural 3D retinal scans. This system uses prior shape knowledge to automatically segment all retinal layers of the 3D-OCT scans using an adaptive, appearance-based method. After the segmentation step, novel texture features are extracted from the segmented layers of the OCT B-scans volume for DR diagnosis. For every layer, Markov-Gibbs random field (MGRF) model is used to extract the 2nd-order reflectivity. In order to represent the extracted imagederived features, cumulative distribution function (CDF) descriptors is used. For layer-wise classification in 3D volume, using the extracted Gibbs energy feature, an artificial neural network (ANN) is fed the extracted feature for every layer. Finally, the classification outputs for all twelve layers are fused using a majority voting schema for global subject diagnosis. A cohort of 188 3DOCT subjects are used for system evaluation using different k-fold validation techniques and different validation metrics. Accuracy of 90.56%, 93.11%, and 96.88% are achieved using 4-, 5-, and 10-fold cross validation, respectively. Additional comparison with deep learning networks which is the state-of-the-art documented the promise of our system’s ability to diagnose the DR early.