Intelligent Multi-Stage Malaria Detection and Life-Stage Classification Using Deep Learning

Tanmay Patil; Mayank Kothari

doi:10.69968/ijisem.2026v5i2145-151

Authors

Tanmay Patil Department of AIML, Mukesh Patel School of Technology, Management and Engineering, Shirpur, India
Mayank Kothari Department of AIML, Mukesh Patel School of Technology, Management and Engineering, Shirpur, India

DOI:

https://doi.org/10.69968/ijisem.2026v5i2145-151

Keywords:

Medical Image Analysis, Malaria Detection, Convolutional Neural Networks , Multi-stage Classification, Transfer Learning

Abstract

The labor-intensive nature of blood smear testing for the presence of malaria in humans requires a trained individual in order to yield accurate test results so to reduce the lengthy testing process, we propose a two-tiered deep learning method for both diagnosing individuals for the presence of malaria and classifying their infection(s) by the stage of the parasite (i.e. ring, trophozoite, schizont, or gametocyte). The first tier will utilize a convolutional neural network (CNN) to make the determination of whether an image has evidence of a malaria infection. The second tier will classify those images that contain malaria parasites based on the parasite life stage. To improve the interpretability of the results, we will use Gradient-weighted Class Activation Mapping algorithm (Grad-CAM) allowing for more accurate visualization of the reasoning behind the way the model classified each image and establish a confidence threshold so that experts may investigate the classification for confirmation. We have conducted preliminary testing on the system for its ability to detect malaria and the preliminary results have been favorable but the development work needed to classify the different life stages of the malaria parasite is still underway. This system provides the ability to rapidly, accurately and interpretably identify malaria. In addition to providing a quick method for diagnosing malaria, this system may be applied to low-resource settings depending on available resources. Lastly, the implementation of data augmentation and preprocessing techniques will increase the performance of the model. Taking into consideration that the design of the system is modular, there are no limitations in developing additional models for future integration into this system. This research highlights the capabilities of integrating deep learning with explainability techniques to better inform the physician or clinician based on medical images.

References

[1] World Health Organization, World Malaria Report 2023. Geneva, Switzerland: WHO, 2023.

[2] M. Mujahid, F. Rustam, R. Shafique, et al., "Efficient deep learningbased approach for malaria detection using red blood cell images," IEEE Access, vol. 12, pp. 12345-12358, 2024.

https://doi.org/10.1038/s41598-024-63831-0

[3] Q. A. Arshad, M. Ali, S. Hassan, C. Chen, et al., "A benchmark dataset and deep learning framework for malaria parasite life-cycle

classification," Computers in Biology and Medicine, vol. 152, 2023.

https://doi.org/10.1016/j.compbiomed.2022.106418

[4] R. Sivaramakrishnan, S. Antani, M. Jaeger, et al., "Pre-trained convolutional neural networks as feature extractors toward improved malaria parasite detection," PeerJ, vol. 8, 2020.

[5] S. Rajaraman, S. Candemir, G. Thoma, and S. Antani, "Visualizing and explaining deep learning predictions for malaria diagnosis," IEEE Journal of Biomedical and Health Informatics, 2021.

[6] A. Abbas, M. M. Abdelsamea, and M. M. Gaber, "Classification of COVID-19 in chest X-ray images using DeTraC deep convolutional

neural network," Applied Intelligence, 2021. (Method reference for CNN generalization)

[7] H. Yang et al., "Deep learning for automated malaria parasite detection and classification," IEEE Access, 2021.

[8] T. Rahman et al., "Transfer learning with deep CNN for malaria parasite detection," Sensors, 2021.

[9] K. Bibin, M. S. Nair, and P. Punitha, "Malaria parasite detection from peripheral blood smear images using deep belief networks," IEEE Access, 2020.

[10] S. Fuhad et al., "Deep learning based automatic malaria parasite detection from blood smear and its smartphone-based application," Diagnostics, 2020.

https://doi.org/10.3390/diagnostics10050329

[11] J. Liang et al., "A deep learning model for malaria classification using microscopic images," BMC Bioinformatics, 2022.

[12] P. Das et al., "Machine learning approach for automated malaria detection using blood smear images," Healthcare Technology Letters, 2022.

[13] Y. Dong et al., "Attention-based deep learning model for malaria parasite classification," Artificial Intelligence in Medicine, 2023.

[14] N. Srivastava et al., "Explainable AI for malaria diagnosis using Grad-CAM and CNN," IEEE Access, 2024.

[15] R. R. Selvaraju, M. Cogswell, A. Das, D. Batra, D. Parikh, and A. Das, "Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization," International Journal of Computer Vision, 2020.

https://doi.org/10.1007/s11263-019-01228-7

[16] Y. Gal and Z. Ghahramani, "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning," in Proceedings of ICML, 2016.