Analysis on retinal diseases using machine learning algorithms

Abstract

The macula is the crucial area of the retina and is of specific interest to retina experts. For the people who distress from diabetes, the existence of anomalies in macula is detected in advance. Otsu’s thresholding algorithm for segmentation is used to categorize the numerous amounts of the body comprising the primary part of the eye called retina are exaggerated. This retinal may prime to swell the surrounding tissue, including the macula. Macular edema ensues when there is anomalous leakage and amassing of fluid in the macula from damaged blood vessels in the nearby retina. The Diabetic Macular Lump (DML) affects patients’ vision that will cause losses in vision. This is astounded by identifying the causes for swelling in an advance stage. The prime causes for the lump are different irregularities in all the blood vessels adjoining the macula. The intention of this effort is to get away from the vision loss by noticing pathologies present in the metaphors of the people exaggerated by macular edema. To classify the images whether they are normal or affected, decision tree classifier is used. The parameters for evaluating performance like accuracy, sensitivity and specificity are found to be 92%, 89.8% and 94.8% respectively.