Scholarlink

Background <p>Effective surveillance of mosquito vectors remains a major challenge in the prevention and control of mosquito-borne diseases such as dengue and malaria. Drone, when integrated with machine learning (ML), offer new opportunities for automated detection and characterization of mosquito breeding habitats. However, evidence on their application across mosquito vector surveillance remains fragmented.</p> Objectives <p>This systematic review aimed to synthesize current evidence on drone-based machine learning applications for mosquito vector surveillance, with particular emphasis on Aedes breeding habitat detection, while incorporating methodologically relevant studies involving other mosquito genera.</p> Methods <p>A systematic search was conducted across Web of Science, Scopus, PubMed, and IEEE Xplore (2000–2024), following PRISMA guidelines. Eligible studies employed drone imagery integrated with ML approaches for mosquito vector surveillance, including <i>Aedes</i> and <i>Anopheles</i>. Data were extracted on study objectives, ML approaches, imagery characteristics, validation strategies, and performance outcomes. Methodological quality was assessed using an adapted MMAT framework.</p> Results <p>Eight studies met the inclusion criteria. Most focused on breeding habitat detection, predominantly targeting <i>Aedes</i> mosquitoes, while a smaller subset addressed Anopheles larval habitats. Deep learning approaches were most commonly applied to detection-oriented tasks, whereas traditional machine-learning methods were primarily used for ecological and habitat characterization, reflecting task-specific methodological preferences across studies.</p> Conclusions <p>Drone-based ML approaches show strong potential for advancing mosquito vector surveillance, particularly for automated habitat detection. While <i>Aedes</i>-focused applications dominate, methodological insights from <i>Anopheles</i> studies contribute to broader vector surveillance strategies. Clear differentiation between empirical evidence and conceptual extensions is essential for guiding future operational deployment.</p> Clinical trial registration <p>Not applicable.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Drone-based machine learning approaches for mosquito vector surveillance

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月20日

查看全文
  • Zulfadli Mahfodz,
  • Agus Naba,
  • Pradeep Isawasan,
  • Ahmad Mohiddin Mohd Ngesom,
  • Nazri Che Dom

摘要

Background

Effective surveillance of mosquito vectors remains a major challenge in the prevention and control of mosquito-borne diseases such as dengue and malaria. Drone, when integrated with machine learning (ML), offer new opportunities for automated detection and characterization of mosquito breeding habitats. However, evidence on their application across mosquito vector surveillance remains fragmented.

Objectives

This systematic review aimed to synthesize current evidence on drone-based machine learning applications for mosquito vector surveillance, with particular emphasis on Aedes breeding habitat detection, while incorporating methodologically relevant studies involving other mosquito genera.

Methods

A systematic search was conducted across Web of Science, Scopus, PubMed, and IEEE Xplore (2000–2024), following PRISMA guidelines. Eligible studies employed drone imagery integrated with ML approaches for mosquito vector surveillance, including Aedes and Anopheles. Data were extracted on study objectives, ML approaches, imagery characteristics, validation strategies, and performance outcomes. Methodological quality was assessed using an adapted MMAT framework.

Results

Eight studies met the inclusion criteria. Most focused on breeding habitat detection, predominantly targeting Aedes mosquitoes, while a smaller subset addressed Anopheles larval habitats. Deep learning approaches were most commonly applied to detection-oriented tasks, whereas traditional machine-learning methods were primarily used for ecological and habitat characterization, reflecting task-specific methodological preferences across studies.

Conclusions

Drone-based ML approaches show strong potential for advancing mosquito vector surveillance, particularly for automated habitat detection. While Aedes-focused applications dominate, methodological insights from Anopheles studies contribute to broader vector surveillance strategies. Clear differentiation between empirical evidence and conceptual extensions is essential for guiding future operational deployment.

Clinical trial registration

Not applicable.