A Comparative Analysis by KNN, SVM & ELM Classification to Detect Sickle Cell Anemia

Abstract

Red blood cell abnormalities involve erythrocytes that supply oxygen to all body tissues. Sometimes the formation and role of erythrocytes are hindered. Sickle cell anemia (SCA) is one kind of red blood cell disease. People carrying sickle cell anemia are increasing day by day. Sickle cell anemia shortens life expectancy. But life expectancy can be extended by diagnosing it an early stage. To identify the existence of sickle cells, an image processing procedure is developed. Blood samples are collected in the form of image format. The conversion of gray image, noise filtering and enhancement of image is done in image pre-processing. Fuzzy C means clustering is applied to determine the normal and sickle cells. Morphological operations are also applied to images. The geometrical and statistical features are used for extraction. Lastly, k nearest neighbor (knn), support vector machine (svm) & extreme learning machine (elm) classifiers are implemented to test images. Comparisons among the classifiers with reliable results are presented by this system.