The Leukemia Healthy and Unhealthy Detection with Wavelet Transform Based On Co-Occurrence Matrix and Support Vector Machine

Abstract

Leukemia is a malignant disease and belongs in a broader sense to Cancers. There are many types of leukemia, each of which requires specific treatment. Leukemia is almost one-third of all cancer deaths in children and young people. The most common type of leukemia in children is acute lymphoblastic leukemia (ALL). In this paper, a new approach is implanted on Leukemia ALL database. For the method the wavelet transform is used for feature extraction, the gray level co-occurrence matrix is used. Also, for classification, the SVM (Support Vector Machine) method is used. The proposed method is the best in applying the system designed to the Local Binary Pattern (LBP) and Histogram of Orientation (HOG) methods. This system aims to detect, diagnose, and verify leukemia cells from microscopic images to get high accuracy, efficiency, reliability, less processing time, smaller error, not complexity, fast, and easy to work. The system was built using microscopic images by examining changes in texture, colors, and statistical analysis. The success rate was 96.1667% for cancer data and 99.8833% for non-cancer data.