
A paper of 4 February 2026 in PLOS ONE evaluated larval source management (LSM) using drone imagery and an AI model that ranks the risk of water bodies, in eight sub-districts of Kwaebibirem District, Eastern Region, Ghana: identified breeding sites rose more than threefold, larvicide use fell by over 60% and labour by about 50%. Malaria case trends were comparable with control sub-districts.
Main points
- Journal: PLoS One 21(2): e0340690. DOI 10.1371/journal.pone.0340690. Received 14 June 2025; accepted 24 December 2025; published 4 February 2026. Authors: Bokpin GA, Adzei FA, Dadzie S, Umeda M, Kim J.
- Workflow: a DJI Mavic 3 Multispectral drone (RGB camera only) mapped the area 1–2 days before treatment; AI detected water bodies and labelled them high or low risk; coordinates were sent to spray teams through a mobile app.
- Design: four intervention sub-districts (Abaam, Asuom, Takyiman, Nkwantanang) and four control sub-districts with manual LSM (Subi, Kade, Pramkese, Abodom). The intervention ran in the dry season, September–December 2024, about four weeks per sub-district. The AI model was trained on 4,021 water bodies surveyed in January–August 2024 in Sierra Leone and Ghana.
- Limitations noted by the authors: dry season only, with below-average rainfall; longer evaluation across transmission seasons is needed.
VPMA’s note
A trial against malaria mosquitoes in Ghana. The “survey first, treat the right spots” approach is useful when members treat larvae over large areas; it does not show an epidemiological effect.
- Source
- PLOS ONE
- Published
- 4 February 2026
- Section
- New approaches to insect control
- Journal
- PLoS One 21(2): e0340690
- DOI
- 10.1371/journal.pone.0340690
Editorial note: This is a short summary and comment by VPMA, not a full copy. The original text and copyright belong to the source cited. Picture: Figure 3 of the PLOS ONE paper (water bodies found with drone/AI and with the conventional method in four sub-districts, CC BY licence).
Source: PLOS ONE (summarised in English by VPMA). Read the Vietnamese page.