Distribution of retinal images with optic diskin proposed hybrid approaches

Medical imaging plays an important role in diagnosis and treatment by ensuring that the internal structure of the human body can be seen with various imaging tools. Detection and diagnosis of diseases in retinal images can be done more quickly with developed computer technologies. Early diagnosis and treatment of retinal eye diseases is important to prevent vision loss. Vascular structures on the retinal image, optic disc, etc. the location of the morphological structures and the deterioration of these structures are very important for the diagnosis and follow-up of the disease. Faults can be detected due to physical reasons such as fatigue when physicians identify abnormalities on the optical disc. Automatic systems were needed to reduce the error rate and the workload of the physicians, and to speed up the identification of parameters to diagnose the disease. Visual examination and analysis of large data sets brings with it a huge workload especially for healthcare workers. Considering this situation, the development and use of fully automatic systems will reduce this load considerably. In this study, semi-automatic segmentation methods are combined with localization method for optical disk segmentation and fully automated segmentation is performed. The performance of the end result methods in optical disk segmentation is determined and evaluated. For this purpose, a hybrid method is proposed for the determination and segmentation of the optical dyskin location in retinal images by mentioning the L1 minimization algorithm, active contour method and region growing algorithm from the morphological operations. The hybrid method developed for optical disc detection has been tested on many images obtained from Fırat University Ophthalmology Department. very promising results have been achieved.