Detection of Acute Lymphoblastic Leukemia Using Random Forest Model

This study suggests a novel technique for the diagnosis of acute lymphoblastic leukemia (ALL) using peripheral blood smear (PBS) images using a random forest machine learning model. Leukemia is a hematological cancer characterized by excessively high immature blood cell proliferation. Acute lymphoblastic leukemia is a fast-moving kind of leukemia in which healthy bone marrow cells that produce functional lymphocytes are replaced by malignant cells that cannot fully develop. The circulation carries the leukemia cells to many organs and tissues, including the brain, liver, lymph nodes, and testes, where they continue to proliferate. The standard technique of diagnosing ALL involves a combination of microscopic inspection of bone marrow and blood samples, laboratory testing, and clinical assessment. Apart from being time-consuming, subjective, and heavily dependent on the competency of the medical professionals conducting and interpreting the tests, these processes also need help to process and evaluate large amounts of data rapidly. In the following study, we leverage the power of ensemble learning by combining multiple decision trees that have been trained on a large dataset. The study uses characteristics extracted from PBS images to identify important factors indicative of leukemia, including abnormal lymphoblast size, shape, and quantity as well as the presence of immature white blood cells and lymphoblasts. With the help of our random forest, we can distinguish between samples that have leukemia and those that are normal. The results show encouraging specificity, sensitivity, and accuracy, indicating that this strategy may work well for automated leukemia diagnosis in medical imaging applications.