An Effective WBC Segmentation and Classification Using MobilenetV3–ShufflenetV2 Based Deep Learning Framework

Abstract

White Blood Cells are essential in keeping track of a person's health. However, the pathologist's experience will determine how the blood smear is evaluated. Furthermore, it is still challenging to classify WBCs accurately because they have various forms, sizes, and colors due to distinct cell subtypes and labeling methods. As a result, a powerful deep learning system for WBC categorization based on MobilenetV3-ShufflenetV2 is described in this research. Initially, the WBC images are segmented using an efficient Pyramid Scene Parsing Network (PSPNet). Following that, MobilenetV3 and an Artificial Gravitational Cuckoo Search (AGCS)-based technique are used to extract and select the global and local features from the segmented images. Finally, the WBC images are divided into five classes using a ShufflenetV2 model. The proposed approach is evaluated on blood cell count and detection (BCCD) and Raabin-Wbc datasets and achieves 99.19% and 99% accuracy, respectively. Moreover, the results are satisfactory when compared to existing algorithms. INDEX TERMS white blood cells, deep learning, MobilenetV3, ShufflenetV2

Blood cell image segmentation and classification: a systematic review

“…This issue is consistently highlighted as a significant challenge in developing and validating new models. Another common challenge mentioned in the surveys ( Das et al, 2022 ; Khan et al, 2021 ; Anilkumar, Manoj & Sagi, 2023 ; Raina et al, 2023 ; Rao & Rao, 2023 ; Umamaheswari & Geetha, 2019 ; More & Sugandhi, 2023 ; Al-Dulaimi & Makki, 2023 ; Veeraiah, Alotaibi & Subahi, 2023 ) is the inherent complexity in microscopic images. This shows a demanding need for robust algorithms that can handle such complexities.…”

“…This shows a demanding need for robust algorithms that can handle such complexities. Additionally, the inconsistency in diagnosis depending on the pathologist's experience is emphasized in surveys ( Das et al, 2022 ; Raina et al, 2023 ; Rao & Rao, 2023 ; Umamaheswari & Geetha, 2019 ; More & Sugandhi, 2023 ; Veeraiah, Alotaibi & Subahi, 2023 ), highlighting the importance of automated systems for more consistent and accurate diagnosis. Each survey also brings unique contributions like ( SivaRao & Rao, 2023 ) and ( Bhargavi et al, 2023 ), which uniquely focus on using SegNet, EfficientNet, and XGBoost for WBC classification, providing a different perspective from other surveys.…”

“…The survey in Veeraiah, Alotaibi & Subahi (2023) focuses on a technical article that proposes a new method, the Histogram Threshold Segmentation Classifier (HTsC), rather than a traditional survey. Regarding weaknesses, some of the previous works have some common limitations, including ( Khan et al, 2021 ; Thomas & Sreejith, 2018 ; Byndur et al, 2023 ; Rao & Rao, 2023 ; Umamaheswari & Geetha, 2019 ; Al-Dulaimi & Makki, 2023 ; Asghar et al, 2023 ), the absence of experimental results or comprehensive comparisons between methods. In contrast, the proposed survey appears to be more comprehensive and holistic.…”

Using deep DenseNet with cyclical learning rate to classify leukocytes for leukemia identification

“…Another a hybrid approach of recurrent neural networks (RNNs). Leukocyte segmentation was implemented using a network based on W-Net, a CNN-based technique for WBC classification implemented by Rao and Rao ( 24 ). Afterward, a DL system based on GhostNet was used to retrieve important feature maps.…”

“…The proposed method has attained an ACC of 99.24% on the Blood Cell Count and Detection (BCCD). Rao and Rao ( 24 ) presented another DL-based framework for the classification of leukocytes based on the MobilenetV3-ShufflenetV2 DL paradigm. At first, an effective Pyramid Scene Parsing Network (PSPNet) is used to segment the images.…”