A computer-assisted human peripheral blood leukocyte image classification method based on Siamese network

Wang, Yapin; Cao, Yiping

doi:10.1007/s11517-020-02180-2

Cited by 8 publications

(4 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our study achieved an accuracy of 97.7% for white blood cell classi cation which is better than some of the reported accuracies 96.1% and 96.9% respectively [10,11]. Our overall accuracy is comparable with the reported average accuracy of 98.8% in a recent study that used Siamese network [35], however, that study was restricted to white blood cells, and it used only 430 images for testing and moreover, no visual saliency mappings is implemented for explainability. number of images per class in the support set could lead to greater classi cation performance.…”

Section: Discussionsupporting

confidence: 65%

Few-Shot Learning using Siamese Twin Network for the Classification of Blood Cells

Tummala

Suresh

2022

Preprint

View full text Add to dashboard Cite

Automated classification of blood cells from microscopic images is an interesting research area owing to advancements of efficient neural network architectures. Here, we developed a few-shot contrastive learning model for the classification of peripheral blood cells including lymphocytes, monocytes, basophils, eosinophils, neutrophils, immature granulocytes, erythroblasts, and platelets using EfficientNet as a base model and contrastive loss as a loss function. A total of 17092 publicly accessible images acquired using the CellaVision DM96 were analyzed. From 125 images of each cell type, 20000 image pairs are generated for Siamese twin network (STN) training and another 125 images from each cell type are used for few-shot validation. Therefore, out of 17092 images, 6% were used for training, 6% for few-shot validation and rest 88% for few-shot testing. This architecture demonstrates an overall accuracy of 97.21% during 8-way 3-shot testing for the classification of all cell types with an accuracy of 97.72% for the classification of white blood cells alone. Further, we propose a novel class activation mapping scheme for the interpretability of the model decisions suitable for STN. To conclude, the proposed framework based on contrastive learning could be used for the fully automated self-exploratory classification and identification of peripheral blood cells.

show abstract

Section: Discussionsupporting

confidence: 65%

Few-Shot Learning using Siamese Twin Network for the Classification of Blood Cells

Tummala

Suresh

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Since the number of positive case pairs is rare and the category is unbalanced, the author uses data amplification to increase the number of positive case pairs. Experimental results show that the siamese structure has advantages for small datasets [ 12 ]. Min Liu proposes an improved autoencoder (AE) network, which uses siamese framework and Gaussian pyramid for multi-scale processing of input images.…”

Section: Related Workmentioning

confidence: 99%

“…CNN models have been successfully applied to CT image recognition of brain diseases [ 9 ], detection of lung cancer cells [ 10 ], classification of radiology imaging, cardiology imaging, and gastroenterology imaging [ 11 ]. For the second issue, Some scholars have used siamese networks and data augmentation [ 12 ] to solve the problem of data scarity. In addition, pretrained models [ 13 ] are also used to solve this problem.…”

Section: Introductionmentioning

confidence: 99%

Automated fundus ultrasound image classification based on siamese convolutional neural networks with multi-attention

Tan

Dong

2023

BMC Med Imaging

View full text Add to dashboard Cite

Fundus ultrasound image classification is a critical issue in the medical field. Vitreous opacity (VO) and posterior vitreous detachment (PVD) are two common eye diseases, Now, the diagnosis of these two diseases mainly relies on manual identification by doctors. This method has the disadvantages of time-consuming and manual investment, so it is very meaningful to use computer technology to assist doctors in diagnosis. This paper is the first to apply the deep learning model to VO and PVD classification tasks. Convolutional neural network (CNN) is widely used in image classification. Traditional CNN requires a large amount of training data to prevent overfitting, and it is difficult to learn the differences between two kinds of images well. In this paper, we propose an end-to-end siamese convolutional neural network with multi-attention (SVK_MA) for automatic classification of VO and PVD fundus ultrasound images. SVK_MA is a siamese-structure network in which each branch is mainly composed of pretrained VGG16 embedded with multiple attention models. Each image first is normalized, then is sent to SVK_MA to extract features from the normalized images, and finally gets the classification result. Our approach has been validated on the dataset provided by the cooperative hospital. The experimental results show that our approach achieves the accuracy of 0.940, precision of 0.941, recall of 0.940, F1 of 0.939 which are respectively increased by 2.5%, 1.9%, 3.4% and 2.5% compared with the second highest model.

show abstract

“…Recently, the field of computer vision has seen swift advancements in object detection algorithms powered by deep learning convolutional neural networks (CNNs), such as 3D shape reconstruction [2,3], superresolution imaging [4,5], leukocyte classification [6,7] and so on. These algorithms are divided into two categories based on the detection stage: two-stage and single-stage algorithms.…”

Section: Introductionmentioning

confidence: 99%

Enhanced floating debris detection algorithm based on CDW-YOLOv8

Shi,

Lei,

You

et al. 2024

Phys. Scr.

View full text Add to dashboard Cite

The issue of floating debris on water surfaces is becoming increasingly prominent, posing significant threats to aquatic ecosystems and human habitats. The detection of floating debris is impeded by complex backgrounds and water currents, resulting in suboptimal detection accuracy. To enhance detection effectiveness, this study presents a floating debris detection algorithm rooted in CDW-YOLOv8. Firstly, the study augments the original C2f module by incorporating the Coordinate Attention (CA) mechanism, resulting in the C2f-CA module, to boost the model’s sensitivity to target locations. Secondly, the study substitutes the standard Upsample module with the DySample module to diminish model parameters and increase flexibility. Furthermore, the study incorporates a small object detection layer to enhance the detection performance of small floating debris. Lastly, the Complete-IOU (CIOU) loss function is substituted by the Focaler-Wise-IOU v3 (Focaler-WIoUv3) loss function, which aims to minimize the impact of low-quality anchor boxes and improve regression accuracy. Experimental results demonstrate that the improved CDW-YOLOv8 algorithm has realized a comprehensive performance improvement in accuracy, recall rate, mAP@0.5, and mAP@0.5:0.95, noting increases of 2.9%, 0.6%, 2.5%, and 1.5%, respectively, relative to the original YOLOv8 algorithm. This offers a robust reference for the intelligent detection and identification of floating debris on water surfaces.

show abstract

A computer-assisted human peripheral blood leukocyte image classification method based on Siamese network

Cited by 8 publications

References 31 publications

Few-Shot Learning using Siamese Twin Network for the Classification of Blood Cells

Few-Shot Learning using Siamese Twin Network for the Classification of Blood Cells

Automated fundus ultrasound image classification based on siamese convolutional neural networks with multi-attention

Enhanced floating debris detection algorithm based on CDW-YOLOv8

Contact Info

Product

Resources

About