Bone Marrow Cancer Detection From Leukocytes using Neural Networks
Sundari M. Shanmuga, Pankaj Bhambri · 2024
In this ambitious project, the primary goal is to harness the power of cutting-edge neural networks, specifically ResNet50 and MobileNetV2, to discern between cancerous and normal white blood cells. The focus extends to prevalent blood disorders such as acute lymphoblastic leukemia (ALL) and multiple myeloma (MM), both stemming from anomalies in leukocyte development. These critical blood components, originating in the bone marrow, become central players in the realm of blood-related diseases. ALL manifests through the overproduction of lymphocytes, leading to the unbridled proliferation of white blood cells. Conversely, MM takes a distinctive course, causing cancerous cells to accumulate within the bone marrow, disrupting normal cellular function. The microscopic examination of these blood samples presents a formidable challenge, as distinguishing between lymphoblasts and regular white blood cells proves intricate due to morphological similarities. The strategic utilization of sophisticated neural networks like ResNet50 and MobileNetV2 represents a pivotal advancement, aiming to significantly enhance the accuracy of this discrimination process. This innovative approach holds great promise as it paves the way for efficient diagnosis and classification of blood cancers based on intricate cellular characteristics, revolutionizing the landscape of medical analysis and treatment.