چكيده به لاتين
The retina a part of eye that consists of several layers. The main task of the network is to convert light energy into analytical signals for the brain. The most important diseases of the retina are age-related macular degeneration (AMD) and DME. The most important methods for diagnosing these diseases are accurate imaging of the layers and veins in the retina. The most advanced and latest OCT imaging technique is providing ophthalmologists with valuable precision tomographic sections of the network layers. The main purpose of processing network images is to create a system for analyzing these images in order to diagnose diseases and their related complications. Floyd areas in OCTs are one of these complications, which is a good indicator of AMD and DME. Floyd areas are also called cysts. The main focus of this study is to divide the Floyd area into AMD and DME patients. In this project, a conversion is proposed to transfer the image from pixel space to the theoretical space and an automated method for floating in the two diseases in the tropical space using fuzzy clustering. In the fragmentation of the Floyd DME patients, the first and last layers, as well as the two intermediate layers of each OCT scan, are grouped by methods based on the shortest path of the graph by providing methods for calculating the gravity of the graph. The first and last layers are used to separate the foreground from the image and the middle layers for postprocessing. Then, by defining a target function and minimizing it, the clustered image, which leads to the segmentation of the Floyd area. The number of clusters needed for each scan is calculated manually. Ultimately, the hemorrhage between the intermediate layers is eliminated to achieve the final fragmentation of the Floyd for DME patients. In the Floyd section of AMD patients, the first and last layers are first partitioned and the last layer is smooth. Then the primary points of Floyd and texture are determined by an automatic method in the theatrosafic space. In the following, we propose a goal function in the proposed tropical space and minimize this function by methods based on the gradient reduction of the image. Finally, the validity of the various stages of the algorithms is proved and the numerical results of each stage are analyzed. The clinical significance of the Floyd area has also been investigated, and it is shown that these areas are slightly related to the sight of individuals and are important for the diagnosis and treatment of retinal diseases
