ABSTRACT
Computational techniques have great impact in the field of Medicine and Biology. These techniques help the medical practitioners to diagnose any abnormality in advance and provide fruitful treatment. Retinal image analysis has been an ongoing area of research. Automated retinal image analysis aid the ophthalmologists in detecting abnormalities in the retinal structures namely optic disc, blood vessels, thus diagnosing sight threatening retinal diseases such as Glaucoma and Retinopathy. Glaucoma is the major cause of blindness in working population. Glaucoma is characterized by increased intra-ocular pressure inside the eye leading to changes in the optic disc and optic nerve. It does not reveal its symptoms until later stage. Hence, regular screening of the patients is required to identify the disease, thus demanding high labor, time and expertise. Thus, computational techniques are sought for their analysis. In this project, identification of Glaucoma is carried out through computational techniques namely image processing. As the changes in the profile of optic disc act as a biomarker for the onset of the disease, optic disc is segmented through image processing techniques. Optic disc is the brightest part portrayed as oval structure in the retinal fundus image. It encompasses optic cup, which is the brightest central part, optic rim, the surrounding pale part and the blood vessels. All these structures are segmented and their properties are elicited. Then, properties of the disc, cup and blood vessels within optic disc are mined to design a learning model for prediction of Glaucoma.
PROJECT OUTPUT
PROJECT DEMO VIDEO
.png)
