Leukemia Diagnosis Using Image Processing and Computational Intelligence
Hamed Parvaresh, Hedieh Sajedi, Seyed Amirhosein Rahimi · 2018
Acute Lymphoblastic Leukemia (ALL) is the most prevalent acute leukemia in adults after Acute Myeloid Leukemia, with a diffusion of over 6500 persons per year just in the United States. In this research, we propose a smart assistant determination method for ALL diagnosis using microscopic images. In this regard, K-means is employed to extract cell images after that wavelet transform is hired on cell images then statistical moments of the transformed image are computed to extract features. Afterward, a Chain Tabu search algorithm is proposed for feature selection of normal and abnormal cells to enable classifiers classifying ALL efficiently. Finally, Multi-Layer Perceptron (MLP) is used for classification. The proposed method is evaluated on ALL-IDB2. The proposed method achieved the accuracy of 98.88% and outperforms existed ALL diagnosis methods.