A Review of Deep Learning-Based Methods for Cell Image Recognition and Tracking

Qiancheng Zhou

doi:10.54254/2753-8818/2025.AU28768

Theoretical and Natural ScienceOpen access

A Review of Deep Learning-Based Methods for Cell Image Recognition and Tracking

Research Article

Open Access

A Review of Deep Learning-Based Methods for Cell Image Recognition and Tracking

Qiancheng Zhou ^1*

¹ Nanjing University of Posts and Telecommunications

^*Corresponding author: 2084392260@qq.com

Published on 28 October 2025

TNS Vol.147

ISSN (Print): 2753-8826

ISSN (Online): 2753-8818

ISBN (Print): 978-1-80590-489-2

ISBN (Online): 978-1-80590-490-8

Download Cover

Abstract

By enabling precise quantification, cell image analysis advances understanding of cellular processes and structures. However, traditional methods have limitations in terms of accuracy and efficiency for dense target segmentation, continuous trajectory recognition, and large-scale data processing. Thus, this paper examines deep learning approaches for cell recognition and tracking, highlighting advancements in tracking across frames, automatic feature extraction, model adaptability, and their effectiveness to handle diverse and complex cellular environments. Through the review of relevant literature, including convolutional neural networks (CNNs), Mask R-CNN, HOG-SVM, as well as transfer learning methods, the potential applications in real-time processing, multimodal fusion, and high-throughput analysis are discussed. The results demonstrate that deep learning techniques enable precise segmentation, stable cross-frame tracking, and strong feature extraction in complex, dense cellular environments. Unlike traditional algorithms, deep learning methods notably reduce segmentation errors and tracking interruptions, all while maintaining solid generalization with minimal labeled data.

Keywords:

Imaging system, Target detection, Cell segmentation, Convolutional neural network, Three-dimensional tracking

View PDF

References

[1]. Henry, K.M., Pase, L., Ramos-Lopez, C.F., Lieschke, G.J., Renshaw, S.A., & Reyes-Aldasoro, C.C. (2013). PhagoSight: An open-source MATLAB® package for the analysis of fluorescent neutrophil and macrophage migration in a zebrafish model. PLOS ONE, 8(8), e72636.

[2]. Liu, Y., Smith, J., Chen, X., & Johnson, M. (2018). MALDI-MSI of immunotherapy: Mapping the EGFR-targeting antibody cetuximab in 3D colon-cancer cell cultures. Analytical Chemistry, 90(24), 14156-14164.

[3]. Luo, W.C., & Guo, W.B. (2000). A comparison and analysis of image thresholding segmentation methods. Modern Computer, (11), 22-25.

[4]. Zhang, C., Tang, K.L., Zhang, H.Q, & Pan, S.H. (2021). A new image segmentation method based on active contour model. Journal of Chengdu University (Natural Science Edition), 40(01), 48-51.

[5]. Liu, XY., & Wang, Y.M. (2019). Target tracking method based on the improved tracking-learning-detection algorithm using HOG-SVM. Science Technology and Engineering, 19(27), 266-271.

[6]. Ramakrishnan, V., et al. (2024). Nuclei detection and segmentation of histopathological images using a feature pyramidal network variant of a Mask R-CNN. Bioengineering, 11(10), 994.

[7]. Duari, S., Gautam, V., & Ahuja, G. (2025). Protocol for cellular age prediction in yeast and human single cells using transfer learning. STAR Protocols, 6(3), 104023.

[8]. Sun, Y. (2022). Research on the identification of suspended cells and multi-cell tracking in microscopic images. Beijing University of Chemical Technology.

[9]. Aleynick, N., Li, Y., Xie, Y., et al. (2023). Cross-platform dataset of multiplex fluorescent cellular object image annotations. Scientific Data, 10(1), 193.

[10]. Doroshenko, O.V., et al. (2024). Automated assessment of wheat leaf disease spore concentration using a smart microscopy scanning system. Agronomy, 14(9), 1945.

[11]. Wang, Q., Zhao, W., Pang, H., et al. (2025). Discussion on the value of multimodal image fusion technology. Advanced Journal of Nursing, 6(2).

[12]. Wu, D., Tan, X., Wang, H., et al. (2025). DMF-LP: Enhancing image quality through multimodal medical image fusion using diffusion and Laplacian techniques. Biomedical Signal Processing and Control, 106, 107890.