article · 2025

Research on Blood Cell Carcinogenesis Classification Method Based on Lightweight Convolutional Neural Network Models

Cited by 0Open access

Abstract

In cases of surplus amount and morphological variation of blood cells, it will present some symptoms of other blood diseases. Therefore, the identification and classification of blood cell images have become important techniques for treating diseases. With the improvement of the application efficiency of computer vision and deep learning in classifying cancerous blood cells, problems such as small training dataset, class imbalance, cell overlap and staining variation still exist, which cause the high cost of training and low classification accuracy. This paper focuses on solving the key problem of accurate classification of blood cells based on image processing and convolutional neural network (CNN) models. This study has developed the BloodNet-V1 CNN model based on three layers of convolution, max pooling, ReLU activation, Adamax optimization, and cross-entropy loss for training. The testing accuracy reached 99% on the eight-class blood cell dataset, and the weighted average F1 score was 98%. This model effectively avoids the high cost of training and overfitting, and then the validity of the algorithm in practical applications has been verified.

Cite this paper

APA 7

Qi, J. (2025). Research on Blood Cell Carcinogenesis Classification Method Based on Lightweight Convolutional Neural Network Models. https://doi.org/10.1145/3784013.3784085

MLA 9

Qi, Jianing. “Research on Blood Cell Carcinogenesis Classification Method Based on Lightweight Convolutional Neural Network Models.” 2025, pp. 447–51, https://doi.org/10.1145/3784013.3784085.

Chicago

Qi, Jianing. 2025. “Research on Blood Cell Carcinogenesis Classification Method Based on Lightweight Convolutional Neural Network Models.” https://doi.org/10.1145/3784013.3784085.

Harvard

Qi, J. (2025) ‘Research on Blood Cell Carcinogenesis Classification Method Based on Lightweight Convolutional Neural Network Models’. Available at: https://doi.org/10.1145/3784013.3784085.

BibTeX

@article{qi2025research,
  title = {{Research on Blood Cell Carcinogenesis Classification Method Based on Lightweight Convolutional Neural Network Models}},
  author = {Qi, Jianing},
  pages = {447--451},
  year = {2025},
  doi = {10.1145/3784013.3784085}
}

Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.