A Systematic Review of Acute Leukemia Diagnosis by Using Deep Learning
Keywords:Leukemia, Deep Learning, Classification, CNN, Segmentation
Acute leukaemia, a malignancy that starts in the bone marrow, manifests as an unchecked, fast spread of white blood cells. Acute lymphoblastic leukaemia is more common in young kids under the age of 5 and older people over the age of 50. Automated diagnosis systems are increasingly using image analysis and Deep Learning techniques due to their great accuracy in a variety of health diagnostic sectors, such as the detection of Acute Lymphoblastic Leukemia from serum samples. The several automated techniques for identifying and categorising acute leukaemia are discussed in this systematic review research along with the challenges the authors faced. In this review, Sahlol, A.T. et al. employed ALL-IDB2 to performed segmentation using the Zack algorithm and achieved the highest segmentation accuracy (99.23%), sensitivity (100%) and uniqueness (97.1%). The study makes use of a number of methodologies, including thresholding, the Zack algorithm, the clustering-based technique, the fuzzy C-mean, the k-myeloid, and the k-means. In the domain of medical image analysis, deep learning worked remarkably well, and CNN made it extremely straightforward to build an end-to-end network. The numerous steps leading up to classification, including preprocessing, augmentation, segmentation, and feature extraction—all of which are discussed in this study.
How to Cite
This is an open Access Article published by Research Center of Computing & Biomedical Informatics (RCBI), Lahore, Pakistan under CCBY 4.0 International License