Optimized convolutional neural network model for multilevel classification in leukemia diagnosis using Tversky loss

Abstract

Leukemia diagnosis traditionally depends on time-intensive examination of blood cell morphology, a process prone to human error. To address these challenges, this study explores the use of convolutional neural networks (CNNs) optimized with the Tversky loss function for automated, multilevel image classification in leukemia diagnostics. The model was designed to tackle binary classification for distinguishing normal from abnormal cells, and multiclass classification for identifying leukemia subtypes, while addressing the challenges of imbalanced datasets inherent in medical imaging. Trained on publicly available leukemia image datasets, the CNN achieved high accuracy in both tasks, effectively capturing subtle morphological variations critical for precise diagnosis. By incorporating performance metrics such as accuracy, precision, and recall, the study highlights the model’s reliability and robustness across classification tasks. The findings underscore the potential of CNN-based tools in enhancing diagnostic accuracy and efficiency, paving the way for future innovations in leukemia diagnostics and broader medical imaging applications.

Keywords

Multilevel classification / Deep learning / Leukemia / Convolutional neural networks / Medical image analysis / Automated diagnosis

Funding

None.

Conflict of interest

The authors declare that they have no competing interests.

References

[1] Varga D. No-reference image quality assessment with multi-scale orderless pooling of deep features. J Imaging. 2021; 7(7):112. doi: 10.3390/jimaging7070112

[2] Li G, Yu Y. Visual saliency detection based on multiscale deep CNN features. IEEE Transact Image Process. 2016; 25:5012-5024. doi: 10.1109/TIP.2016.2602079

[3] Chen Y, Zheng WS, Lai JH, Yuen PC. An asymmetric distance model for cross-view feature mapping in person reidentification. In: IEEE Transactions on Circuits and Systems for Video Technology. United States: IEEE; 2016. p. 1. doi: 10.1109/TCSVT.2016.2515309

[4] Arivuselvam B, Sudha S. Leukemia classification using the deep learning method of CNN. J Xray Sci Technol. 2022; 30(2):199-210. doi: 10.3233/xst-211055

[5] Kadhim KA, Najjar FH, Waad AA, Al-Kharsan IH, Khudhair ZN, Salim AA. Leukemia classification using a convolutional neural network of AML images. Malays J Fundam Appl Sci. 2023; 19(3):560-570. doi: 10.11113/mjfas.v19n3.2901

[6] Talaat FM, Gamel SA. Machine learning in detection and classification of leukemia using C-NMC_Leukemia. Multimed Tools Appl. 2023; 83:8063-8076. doi: 10.1007/s11042-023-15923-8

[7] Rodrigues LF, Backes A, Travençolo B, De Oliveira GD. Optimizing a deep residual neural network with genetic algorithm for acute lymphoblastic leukemia classification. J Dig Imaging. 2022; 35(2):425-435. doi: 10.1007/s10278-022-00600-3

[8] Mallick P, Mohapatra SK, Chae G, Mohanty M. Convergent learning-based model for leukemia classification from gene expression. Pers Ubiquitous Comput. 2023; 25(5):897-906. doi: 10.1007/s00779-020-01467-3

[9] Arif R, Akbar S, Farooq A, Hassan SA, Gull S. Automatic detection of leukemia through convolutional neural network. In: International Conference on Frontiers of Information Technology Proceedings; 2022. doi: 10.1109/FIT57066.2022.00044

[10] Alsaykhan LK, Maashi MS. A hybrid detection model for acute lymphocytic leukemia using support vector machine and particle swarm optimization (SVM-PSO). Sci Rep. 2024; 14:23483. doi: 10.1038/s41598-024-74889-1

[11] Abhishek A, Deb SD, Jha RK, Sinha R, Jha K. Ensemble learning using Gompertz function for leukemia classification. Biomed Signal Process Control. 2025; 100:106925. doi: 10.1016/j.bspc.2024.106925

[12] Loey M, Naman M, Zayed H. Deep transfer learning in diagnosing leukemia in blood cells. Computers. 2020; 9(2):29. doi: 10.3390/computers9020029

[13] Damit DSA, Sulaiman SN, Osman MK, Karim NKA, Razali NF, Marzuki MIF. Navigating Tversky loss function hyperparameter spaces using particle swarm optimization for myocardial scar segmentation. In: 2024 20^th IEEE International Colloquium on Signal Processing and Its Applications (CSPA); 2024. p. 173-177. doi: 10.1109/CSPA60979.2024.10525595

[14] Kumar I, Rawat J. Segmentation and classification of white blood smear images using modified CNN architecture. Discov Appl Sci. 2024; 6:587. doi: 10.1007/s42452-024-06139-y

[15] Muhsen IN, Shyr D, Sung AD, Hashmi SK. Machine learning applications in the diagnosis of benign and malignant hematological diseases. Clin Hematol Int. 2020; 3(1):13-20. doi: 10.2991/chi.k.201130.001

[16] Hehr M, Sadafi A, Matek C, et al. Explainable AI identifies diagnostic cells of genetic AML subtypes. PLOS Digital Health. 2023; 2(3):e0000187. doi: 10.1371/journal.pdig.0000187

[17] Antunes A, Ferreira B, Marques N, Carriço N. Hyperparameter optimization of a convolutional neural network model for pipe burst location in water distribution networks. J Imaging. 2023; 9:68. doi: 10.3390/jimaging9030068

[18] Rodrigues V, Deusdado S. Metalearning approach for leukemia informative genes prioritization. J Integr Bioinformatics. 2020; 17(1):20190069. doi: 10.1515/jib-2019-0069

[19] Bai Y, Yang K, Yu W, Xu C, Zhao T, Ma WY. Automatic image dataset construction from click-through logs using deep neural network. In: Proceedings of ACM SIGIR Conference; 2015. p. 333-342. doi: 10.1145/2733373.2806243

[20] Islam MM, Rifat HR, Shahid MSB, Akhter A, Uddin MA. Utilizing deep feature fusion for automatic leukemia classification: An IoMT-enabled deep learning framework. Sensors. 2024; 24(13):4420. doi: 10.3390/s24134420

[21] Sampathila N, Chadaga K, Goswami N, et al. Customized deep learning classifier for detection of acute lymphoblastic leukemia using blood smear images. Healthcare (Basel). 2022; 10(10):1812. doi: 10.3390/healthcare10101812

[22] Gupta R, Gehlot S, Gupta A. (C-NMC): B-lineage acute lymphoblastic leukaemia: A blood cancer dataset. Med Eng Phys. 2022; 103:103793. doi: 10.1016/j.medengphy.2022.103793