Automatic detection of white blood cancer from bone marrow microscopic images using Convolutional Neural Networks

Over the past few years, computer-aided diagnosis (CAD) has been rapidly advancing, with numerous machine learning algorithms being developed to identify various diseases, including leukemia, a variety of cancer that damages the white blood cells (WBCs) and stays in the bone marrow and/or blood. Early leukemia detection is essential for effective therapy and patient lifesaving. The illness can be categorized into two basic forms—chronic and acute—creating a sum of four subtypes. The proposed study focuses on the two subtypes, Acute Lymphoblastic Leukemia (ALL), Multiple Myeloma (MM), utilizing the SN-AM dataset. ALL happens when it makes many lymphocytes, whereas MM results from an accumulation of cancer cells that stop the bone marrow from adding the healthy blood cells. Normally, defining leukemia subtypes was a time-consuming and error-prone manual process undertaken by qualified professionals. The suggested model, The processes of extracting the precise information from images of cells before identifying the cancer subtype using different algorithms and its techniques, notably CNN. The algorithm outperformed existing ML techniques. With an accuracy rate of 97.2%. The DCNN model's efficiency in identifying the cancer type in the bone marrow was demonstrated by the fact that it performed on par with well-known CNN architectures. The study's overall conclusions show the potential of CAD and deep learning methods in enhancing medical diagnosis and improving patient outcomes.