White blood cells (WBCs) are a vital part of the immune system that protect the body from different types of bacteria and viruses. Abnormal cell growth destroys the body’s immune system, and computerized methods play a vital role in detecting abnormalities at the initial stage. In this research, a deep learning technique is proposed for the detection of leukemia. The proposed methodology consists of three phases. Phase I uses an open neural network exchange (ONNX) and YOLOv2 to localize WBCs. The localized images are passed to Phase II, in which 3D-segmentation is performed using deeplabv3 as a base network of the pre-trained Xception model. The segmented images are used in Phase III, in which features are extracted using the darknet-53 model and optimized using Bhattacharyya separately criteria to classify WBCs. The proposed methodology is validated on three publically available benchmark datasets, namely ALL-IDB1, ALL-IDB2, and LISC, in terms of different metrics, such as precision, accuracy, sensitivity, and dice scores. The results of the proposed method are comparable to those of recent existing methodologies, thus proving its effectiveness.
Blood is a fluid that transports oxygen, providing energy to body cells that then produce carbon dioxide. It also plays a pivotal role in the immune system; blood circulating in living organisms contains 55% plasma, 40% red cells, 4% platelets, and 1% white blood cells (WBCs) [ The Open Neural Network Exchange (ONNX) is applied with a YOLOv2 model, which detects the different types of WBCs. The features are extracted using activation-5 of the ONNX model. The extracted features are fed to the YOLOv2 model. The proposed framework accurately detects the region of interest (ROI). The features are extracted using darknet-53, and the prominent features are selected based on Bhattacharyya separately criteria and fed to the shallow classifiers for the classification of WBCs.
In the literature, significant work has been done for the detection of WBCs, and some of the recent works are discussed in this section [
The proposed approach comprises localization, segmentation, high-level feature extraction/selection, and classification steps for the analysis of WBCs. In the proposed approach, WBCs are detected/localized using ONNX as the backbone of YOLOv2. The localized cells are segmented using the proposed 3-D semantic segmentation model. Finally, the WBCs are classified using multi-SVM. An overview of the proposed method is presented in
In this research, WBCs are recognized by the suggested WBC-ONNX-YOLOv2 model, as shown in
The layer-wise proposed model architecture is presented in
Layers of the proposed model | Activation units | Layers of the proposed model | Activation units |
---|---|---|---|
Image input | |||
Multiple (mul)-mul (element wise affine) | Conv4 | ||
Add-add (element wise affine) | Bn4 | ||
Convolutional (conv) | Act4 | ||
Batch-normalization (Bn) | |||
Activation (act) (leaky ReLU with 0.1 scale) | Conv5 | ||
Bn5 | |||
Conv1 | Act5 | ||
Bn1 | Conv1-YOLOv2 | ||
Act1(leaky ReLU with 0.1 scales) | Bn1-YOLOv2 | ||
ReLU1-YOLOv2 | |||
Conv2 | Conv2-YOLOv2 | ||
Bn2 | Bn2-YOLOv2 | ||
Act2 | ReLU2-YOLOv2 | ||
YOLOv2-classification | |||
Conv3 | YOLOv2-transform | ||
Bn3 | YOLOv2-output | – | |
Act3 |
The proposed model is trained using selected parameters as reported in
Input image size | Training epochs | Batchsize | Optimizers | Learning rate | Training average precision rate (mAP) |
---|---|---|---|---|---|
100 | 12 | Stochastic gradient descent (Sgdm) | 1e −3 | 1.00 |
It is trained on 100 epochs, because after 100 epochs, the model performance is almost stable. The number of iterations with the respective loss during training is illustrated graphically in
The semantic segmentation model is proposed for the segmentation of WBCs, in which deeplabv3 is used as a bottleneck in the Xception model. The pre-trained Xception model contains 205 layers, comprising 1 input, 88 2-D Conv, 46 Bn, 46 ReLU, 3 max-pooling, 12 addition, 4 crop 2D, 2 transpose Conv, 2 depth Conv, softmax, and pixel classification layers. The segmentation model was trained from scratch on the blood smear images. The training parameters of the presented model are listed in
Optimizer | Sgdm |
---|---|
Batch-size | 10 |
Training epochs | 40 |
The proposed model learning with convolutional layers is plotted with activation units, as presented in
The deep features are extracted using a pre-trained darknet53 model, which contains 184 layers, namely 1 input, 53 Conv, 1 global pooling, 52 Bn, 52 LeakyReLU, and 23 addition layers, and softmax with cross-entropy loss. The features are extracted from Conv53 layers with dimensions of
The SVM classifier with different kernels is trained on the best-selected feature vectors with optimum parameters, as listed in
Classifier | Function of the kernel | ||
---|---|---|---|
SVM | Quadratic cubic | Kernel of the scale: automated constraint box level: 1Multilevel approach: One |
|
Optimizable | Kernel scale and box constraint: 0.001–100 |
Types of WBCS | IoU | mAP |
---|---|---|
Acidophilus | 0.95 | 1.00 |
Lymphocytes | 0.92 | 1.00 |
Monocytes | 0.91 | 1.00 |
Basophils | 0.93 | 1.00 |
Neutrophils | 0.90 | 0.80 |
Blast cells | 0.97 | 0.93 |
In this research, three publicly available benchmark datasets are used for the method evaluation. ALL-IDB1 contains 107 blood smear images, of which 33 are blasts and 74 are non-blast cells, and ALL-IDB2 contains 260 blood smear images, comprising 130 blasts and 130 non-blast cells [
The proposed work performance is validated by performing three experiments. The first experiment is performed to validate the presented localization technique by different metrics such as mean precision (mAP) and intersection over the union (IoU). The second experiment is validated to compute the segmentation model performance, while the third experiment is performed to compute the classification model performance. All experiments in this research are performed on the MATLAB 2020 Ra toolbox with 1050 K Nvidia Graphic Card.
Experiment 1 was performed to validate the performance of the localization approach on three benchmark datasets, LISC, ALL-IDB1, and ALL-IDB2, using IoU and mAP as metrics, as shown in
The localization outcomes in
The proposed method localizes the WBCs with confidence scores, as shown in
The localization results in
In this experiment, the 3D segmented region is validated using different types of performance metrics, namely IoU, mean, weighted, and global accuracy, and F1-scores, as mentioned in
Accuracy (global) | Accuracy (mean) | IoU (mean) | IoU (weighted) | F1-score |
---|---|---|---|---|
0.99 | 0.98 | 0.97 | 0.98 | 1.0 |
The segmentation results in
In this experiment, an optimized feature vector is fed to a multi-kernel SVM for WBC classification, and the outcomes are computed in terms of accuracy, precision, recall, and F1 scores from the LISC dataset, as displayed in
Types of WBCS | ACC (%) | PPV | Sensitivity | F1scores |
---|---|---|---|---|
Eosinophils | 100 | 1.0 | 1.0 | 1.0 |
Basophils | 95.78 | 0.90 | 0.90 | 0.90 |
Lymphocytes | 98.95 | 0.98 | 0.97 | 0.97 |
Neutrophils | 96.26 | 0.90 | 0.92 | 0.91 |
Monocytes | 100 | 1.0 | 1.0 | 1.0 |
Types of WBCS | ACC (%) | PPV | Sensitivity | F1scores |
---|---|---|---|---|
Eosinophils | 100 | 1.0 | 1.0 | 1.0 |
Basophils | 92.06 | 0.87 | 0.89 | 0.88 |
Lymphocytes | 97.56 | 0.99 | 0.94 | 0.96 |
Neutrophils | 92.58 | 0.87 | 0.90 | 0.89 |
Monocytes | 100 | 1.0 | 1.0 | 1.0 |
Types of WBCs | ACC (%) | PPV | Sensitivity | F1scores |
---|---|---|---|---|
Eosinophils | 100 | 1.0 | 1.0 | 1.0 |
Basophils | 98.73 | 0.96 | 0.98 | 0.97 |
Lymphocytes | 99.62 | 0.99 | 0.97 | 0.98 |
Neutrophils | 98.83 | 0.97 | 0.97 | 0.97 |
Monocytes | 100 | 1.0 | 1.0 | 1.0 |
A quantitative analysis is performed using an SVM with three different types of kernels, namely cubic, quadratic, and optimized. The SVM with the optimized kernel achieved a maximum overall accuracy of 98.4%. The classification results are also compared with the latest published work, as shown in
The classification results on the ALL-IDB1&2 datasets are presented in
The classification results of blast/non-blast cells are presented in
Ref | Year | Results (accuracy) |
---|---|---|
[ |
2020 | Lymphocyte = 0.995 |
Monocyte = 0.984 | ||
Basophil = 0.984 | ||
Eosinophil = 0.961 | ||
Neutrophil = 0.950 | ||
Lymphocyte = 0.996 | ||
Monocyte = 1.00 | ||
Basophil = 0.987 | ||
Eosinophil = 1.00 | ||
Neutrophil = 0.988 |
Classes | ACC (%) | PPV | Sensitivity | F1scores |
---|---|---|---|---|
Blast cell | 99.57 | 1.0 | 0.99 | 1.0 |
Non-blast cells | 99.57 | 0.99 | 1.0 | 1.0 |
Classes | ACC (%) | PPV | Sensitivity | F1scores |
---|---|---|---|---|
Blast cell | 98.25 | 0.99 | 0.97 | 0.98 |
Non-blast cells | 98.25 | 0.97 | 0.99 | 0.98 |
Ref | Year | Dataset | Results (accuracy) (%) |
---|---|---|---|
[ |
2018 | ALL-IDB | 97.22 |
[ |
2018 | 96.06 | |
[ |
2020 | 97.45 | |
[ |
2020 | 97.00 | |
[ |
2020 | 94.10 | |
Proposed approach | 99.57 |
In this study, deep learning approaches are proposed for the detection of WBCs. Detecting WBCs is challenging because blood smear images contain different color distributions in the cytoplasm and nucleus regions, making it difficult to segment these regions accurately. A 3-D semantic segmentation model is proposed, in which deeplabv3 is used as a bottleneck and the Xception model is used as a classification head to accurately segment the WBCs. Feature extraction/selection is another challenge for the classification of WBCs. The features are extracted from the pre-trained darknet-53 model, and informative features are selected using Bhattacharyya separability criteria and passed to the SVM with different types of kernels, namely cubic, quadratic, and optimized. The proposed classification method achieved an accuracy of 99.57% on the ALL-IDB1 dataset, 98.25% for the ALL-IDB2 dataset, and 98.4% for LISC datasets using the optimizable SVM kernel. The overall experimental outcomes demonstrate that the proposed technique achieved competitive outcomes by optimizing the SVM kernel. The proposed new framework based on a CNN can be used for the detection of different types of cancer, such as lung and bone cancer. It detects and classifies leukocytes at an early stage, thereby increasing the survival rate of patients.
This research was supported by Korea Institute for Advancement of Technology (KIAT).