Leukemia detection using digital image processing in MATLAB

• Main Authors: Dimayuga, Roy A., Ong, Gerwin T., Perez, Rainier Carlo S., Siy, Gefferson O., Sohrabi Langroudi, Saman C.
• Format: text
• Language: English
• Published: Animo Repository 2010
• Subjects: Leukemia; Digital image processing
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/14642
• Institution: De La Salle University

Leukemia is a form of cancer that affects the normal production of blood cells particularly white blood cells or leukocytes. Early detection of the disease is necessary for proper treatment management. The study presents an algorithm for detection and classification of Leukemia through Digital Image Processing. An algorithm that uses various digital image processing techniques capable of extracting morphological features and classifying the blood sample as acute or chronic leukemia or negative for leukemia was devised. The algorithm used features such as size, blood count, and the roundness ratio of the nucleus to determine whether the white blood cell is a blast or a mature leukocyte. Blood count was also considered to compare the concentration of leukocytes. Based on the data gathered, the algorithm was able to achieve 92% accuracy in the detection and classification of Leukemia in 100 blood samples. To get more accurate results and be able to classify specific types of leukemia such as hairy cell leukemia, type of diagnosis such as flow cytometry and immunophenotyping should be explored.