A Deep Learning-based MobileNet Model for Effectively Classifying the White Blood Cancerous Microscopic Images

Traditional machine and deep learning-based techniques effectively utilise advanced medical image analysis (MIA) to improve prediction accuracy and enable optimal planning and diagnosis. This work aims to create an automated classification system using various machine, deep, and ensemble learning-based models. The dataset used in the work contains the B-lineage acute lymphoid leukaemia and Multiple myeloma microscopic images. The models’ performance are evaluated using the Accuracy, precision, recall, specificity, and F1-Score as performance metrics. The findings of the work states that MobileNet is one of the bestperforming models and its accuracy is approx.. 98.04%. A comparative study is also done to prove the efficacy of the model.