Lightweight EfficientNetB3 Model Based on Depthwise Separable Convolutions for Enhancing Classification of Leukemia White Blood Cell Images

Abstract

Acute lymphoblastic leukemia (ALL) occurs when undeveloped lymphocytes are grown excessively in the bone marrow due to the formation of many immature white blood cells (WBC) in the bone marrow to destroy the healthy cell, which may lead to death. White blood cells are part of the immune system fighting against germs. A timely and accurate cancer diagnosis is important for effective treatment to improve survival rates. Since the image of acute lymphoblastic leukemia cells (cancer cells) under the microscope is complicated to recognize the difference between ALL cancer cells and normal cells. In order to reduce the severity of this disease, it is necessary to classify immature cells at an early stage. In recent years, different classification models have been introduced based on machine learning (ML) and deep learning (DL) algorithms, but they need to be improved. This work enhances the diagnosis of ALL with a computer-aided system that yields accurate results by using deep learning techniques. This research proposed a lightweight DL-assisted robust model based on EfficientNet-B3 using depthwise separable convolutions for classifying acute lymphoblastic leukemia and normal cells in the white blood cell images dataset. The main objective of the proposed lightweight EfficientNet-B3 is to utilize less trainable parameters to enhance the performance and efficiency of the classification of leukemia. Furthermore, two publicly available datasets are considered to evaluate the effectiveness and generalization of the proposed lightweight EfficientNet-B3. In addition, different measures are employed to evaluate the effectiveness and efficiency of the proposed classifier for the accurate and reliable classification of leukaemia cells. In addition, a detailed analysis is given to evaluate and compare the performance and efficiency of the proposed with existing pre-trained and ensemble DL classifiers. Based on experimental results, it is investigated that the proposed model for image classification achieves better performance and outperforms the existing benchmark DL and other ensemble classifiers. Moreover, our finding suggests that the proposed lightweight EfficientNet-B3 model is reliable and generalized to facilitate clinical research and practitioners for leukemia detection.