Assessing the impact of data augmentation and a combination of CNNs on leukemia classification

An accurate early-stage leukemia diagnosis plays a critical role in treating and saving patients’ lives. The two primary forms of are acute chronic leukemia, which is subdivided into myeloid lymphoid leukemia. Deep learning models have been increasingly used computer-aided medical (CAD) systems developed to detect This article assesses the impact widely applied techniques, mainly data augmentation multilevel ensemble configurations, deep learning-based CAD systems. Our assessment included five scenarios: three binary classification problems multiclass problems. evaluation was performed using 3,536 images from 18 datasets, it possible conclude that techniques improve performance convolutional neural networks (CNNs). Furthermore, there an improvement results combination CNNs. For problems, configuration superior configuration. However, were statistically similar scenarios. promising, with accuracies 94.73% 94.59% obtained configurations scenario four classes. methods helps reduce error or variance predictions, improves accuracy model.