A crucial step in the diagnosis and treatment of retinal illnesses is the segmentation of the optic disc and exudates in retinal fundus. For precise and effective segmentation of these structures, a hybrid model using a genetic algorithm and a root-guided decision tree is presented in this paper. In retinal imaging, the optic disc is a crucial structural marker, and exudates are important biomarkers for conditions like diabetic retinopathy. The segmentation accuracy is improved by the suggested hybrid model's use of the evolutionary algorithm to optimize decision tree characteristics. Beginning with picture preprocessing, the framework includes color transformation to perceptually uniform color spaces and cropping to establish the field of view. To enhance picture contrast, contrast limited adaptive histogram equalization (CLAHE) is used. Gabor filters are used to improve blood arteries because they successfully capture the Gaussian approximation of these structures. The suggested technique retrieves features for optic disc segmentation include intensity characteristics, neighborhoods statistics, and picture modifications. These characteristics are used to forecast if a pixel is a part of the optic disc-containing area of interest (ROI). The ability to distinguish exudates from the optical disc is aided by post-processing methods and domain expertise. On benchmark datasets like DRIVE and DIARETDB1, the performance of the suggested technique is assessed. The outcomes show excellent performance in both the segmentation of the optic disc and the exudates tests. The competitiveness of the suggested hybrid paradigm is demonstrated through comparison with alternative approaches. The efficacy of the suggested paradigm has important clinical implications. Exudates and the optic disc can be accurately segmented to enable the early identification and monitoring of retinal disorders, facilitating prompt treatment and intervention. Automation of segmentation activities lightens the load on physicians and improves diagnostic effectiveness. In the segmentation of the optic disc and exudates, the hybrid model using a genetic algorithm and root-guided decision tree shows encouraging results. Retinal image analysis is made easier with the help of the suggested method, which also makes it easier to diagnose and treat retinal illnesses.

Optic disc, exudates, retinal images, segmentation, hybrid model, genetic algorithm, root guided decision tree, retinal diseases, diabetic retino pathy, glaucoma, image preprocessing.