A systematic review assessing the role of computer vision in mosquito-borne disease management has revealed that while research has expanded substantially, the majority of developed models lack prospective validation and evaluation in operational settings. Vector surveillance, pathogen detection, and clinical diagnosis represent vital components of disease control, but conventional approaches are constrained by labor-intensive workflows and a lack of trained specialists. Automated visual inspection techniques have therefore emerged across the management pipeline to address these capacity constraints.
Following the PRISMA 2020 guidelines, researchers Song-Quan Ong, Branko Brkljač, Tamara Djerdj, Maria Cholvi, Angeliki F. Martinou, Jessica J. Williams, Domagoj K. Hackenberger, and Branimir K. Hackenberger conducted a comprehensive literature search across PubMed, Scopus, and IEEE Xplore. The team screened 1,521 candidate records published between January 1, 2010, and December 31, 2025. A total of 650 studies met the inclusion criteria and underwent descriptive quantitative analysis and qualitative thematic synthesis.
The review established that diagnostic applications formed the focus of most published studies, particularly the detection of Plasmodium parasites in malaria microscopy. Over the sixteen-year survey window, research activity increased significantly as analytical methods evolved from traditional image processing pipelines toward advanced deep learning systems. Algorithmic strategies transitioned through convolutional neural networks, You Only Look Once (YOLO)-based object detectors, and modern transformer architectures.
More recent investigations expanded beyond microscopic diagnosis into vector ecology and surveillance. Researchers deployed deep learning algorithms for mosquito species classification, utilized unmanned aerial vehicles and satellite imaging for environmental habitat mapping, and developed smartphone-based platforms for non-invasive diagnostic imaging. Furthermore, emerging investigations explored three-dimensional mosquito behavioral analysis alongside surveillance networks enabled by the Internet of Things.
Despite these technological advancements, the authors identified major methodological hurdles that impede translational readiness into public health practice. Most evaluated models were trained and tested on small datasets from limited geographic areas without external validation. Because few algorithms were evaluated within operational field environments, public health translation remains restricted until models demonstrate reliable performance under real-world conditions.
Subject of Research: Applications of computer vision in mosquito-borne disease research and control
Article Title: Applications of computer vision in mosquito-borne diseases: A systematic review
Article References: Ong, S.-Q., Brkljač, B., Djerdj, T., Cholvi, M., Martinou, A. F., Williams, J. J., Hackenberger, D. K., & Hackenberger, B. K. (2026). Applications of computer vision in mosquito-borne diseases: A systematic review. PLOS Neglected Tropical Diseases, 20(11), e0014870. https://doi.org/10.1371/journal.pntd.0014870
Image Credits: AI Generated
DOI: 10.1371/journal.pntd.0014870
Keywords: computer vision, mosquito-borne diseases, systematic review, PRISMA guidelines, malaria microscopy, Plasmodium detection, vector surveillance, pathogen detection, deep learning, convolutional neural networks, YOLO object detectors, transformer architectures, mosquito species classification, habitat mapping, unmanned aerial vehicles, smartphone diagnostic imaging, behavioral analysis, Internet of Things
News Source: Blake Davidson. (November 19, 2026). Systematic Review Finds Computer Vision In Mosquito Control Lacks Operational Testing. Scienmag.



