Automatic detection of acute lymphoblastic leukaemia based on HSI colour space

Image segmentation is the most important step and critical task in image processing. The term leukaemia refers to a group of cancer blood cells which occur when the cells stop working and functioning properly and affect blood forming (Stem) cells in the body. In leukaemia, white blood cells become abnormal, divide and grow in an uncontrolled way. Early detection of the disease is necessary; it aims at aiding the prognosis of the disease through image analysis. It primarily involves noise identification and removal, enhancement and segmentation of blood orbone marrow smear images and analysis of processed images. After effective noise removal, depending upon clarity of image, suitable enhancement technique is employed. Then the image is converted into appropriate colour space and segmentation algorithm is applied to segregate cells from background. This allows the identification of abnormal WBCs. Then feature extraction and analysis of the WBCs was done.