Image-Based WBC Analysis: A Machine Learning Approach for Cancer Detection

Distinguishing between the two states of white blood cells (WBCs, normal and malignant) is important as they are key in the biomedical scenario for the diagnostic treatment of diseases like leukemia made to diagnose ruptures or dysfunctions. In the last decade, with advancement of deep learning and image processing this process has gone through a metamorphosis from being manual (dependent on pathologist expertise) to automated. This paper further investigates the application of convolutional neural networks (CNNs) to automatically classify WBCs, which has gained a lot of importance for early disease detection in developed personalized treatment plans, aging population and increasing laboratory efficiency. Effective clinical application of these systems could yield a very important impact on patients with blood disorders.