Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Nov 30;9(12):266.
doi: 10.3390/jimaging9120266.

YOLO-PAM: Parasite-Attention-Based Model for Efficient Malaria Detection

Luca Zedda  1 , Andrea Loddo  1 , Cecilia Di Ruberto  1
Affiliations

YOLO-PAM: Parasite-Attention-Based Model for Efficient Malaria Detection

Luca Zedda et al. J Imaging. .

Abstract

Malaria is a potentially fatal infectious disease caused by the Plasmodium parasite. The mortality rate can be significantly reduced if the condition is diagnosed and treated early. However, in many underdeveloped countries, the detection of malaria parasites from blood smears is still performed manually by experienced hematologists. This process is time-consuming and error-prone. In recent years, deep-learning-based object-detection methods have shown promising results in automating this task, which is critical to ensure diagnosis and treatment in the shortest possible time. In this paper, we propose a novel Transformer- and attention-based object-detection architecture designed to detect malaria parasites with high efficiency and precision, focusing on detecting several parasite sizes. The proposed method was tested on two public datasets, namely MP-IDB and IML. The evaluation results demonstrated a mean average precision exceeding 83.6% on distinct Plasmodium species within MP-IDB and reaching nearly 60% on IML. These findings underscore the effectiveness of our proposed architecture in automating malaria parasite detection, offering a potential breakthrough in expediting diagnosis and treatment processes.

Keywords: computer vision; deep learning; early malaria diagnosis; image processing; malaria parasite detection.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
Comprehensive overview of the investigated datasets. The figure presents a detailed overview of the two datasets investigated in this study: MP-IDB and IML. MP-IDB encompasses four distinct malaria species—P. falciparum, P. malariae, P. ovale, and P. vivax. In contrast, the IML dataset exclusively consists of samples related to P. vivax. Notably, the MP-IDB dataset demonstrates intra-species variations, while the datasets differ significantly from each other.
Figure 2
Figure 2
Overview of the modules and mechanisms’ hierarchy proposed in this study to enhance the performance of YOLOv8. Here, NAM stands for Normalized Attention Module, while CBAM refers to Convolutional Block Attention Module. Further, C2f is a fast implementation of the Cross-Stage Partial (CSP) Bottleneck with 2 convolutions, while C3 indicates a CSP Bottleneck with 3 convolutions. Finally, C3STR refers to the integration of the Swin Transformer Block in place of the C3 module’s Bottleneck.
Figure 3
Figure 3
The proposed YOLO-PAM architecture.
Figure 4
Figure 4
The middle section shows detection outcomes obtained with the baseline method, YOLOv8m, on a sample image taken from the P.f. split of MP-IDB. A closer look (at the right) reveals missing parasites in the detection, along with the misclassification of a white blood cell as a parasite. In contrast, the lower section presents results obtained with the proposed method, YOLO-PAM. Here, all the parasites are accurately detected, and the white blood cell is not flagged as a parasite. This comparison underscores the enhanced precision and accuracy achieved by YOLO-PAM.
See this image and copyright information in PMC

References

    1. Rantala A.M., Taylor S.M., Trottman P.A., Luntamo M., Mbewe B., Maleta K., Kulmala T., Ashorn P., Meshnick S.R. Comparison of real-time PCR and microscopy for malaria parasite detection in Malawian pregnant women. Malar. J. 2010;9:269. doi: 10.1186/1475-2875-9-269. - DOI - PMC - PubMed
    1. Gimenez A.M., Marques R.F., Regiart M., Bargieri D.Y. Diagnostic methods for non-falciparum malaria. Front. Cell. Infect. Microbiol. 2021;11:681063. doi: 10.3389/fcimb.2021.681063. - DOI - PMC - PubMed
    1. A V., Kanna B.R. Deep learning approach to detect malaria from microscopic images. Multimed. Tools Appl. 2020;79:15297–15317.
    1. Loddo A., Ruberto C.D., Kocher M., Prod’Hom G. MP-IDB: The Malaria Parasite Image Database for Image Processing and Analysis. In: Leporé N., Brieva J., Romero E., Racoceanu D., Joskowicz L., editors. Proceedings of the Processing and Analysis of Biomedical Information—First International SIPAIM Workshop, SaMBa 2018, Held in Conjunction with MICCAI 2018; Granada, Spain. 20 September 2018; Berlin/Heidelberg, Germany: Springer; 2018. pp. 57–65. Revised Selected Papers.
    1. Maity M., Jaiswal A., Gantait K., Chatterjee J., Mukherjee A. Quantification of malaria parasitaemia using trainable semantic segmentation and capsnet. Pattern Recognit. Lett. 2020;138:88–94. doi: 10.1016/j.patrec.2020.07.002. - DOI

LinkOut - more resources