AI-supported automated microscopy for malaria diagnosis
This article has been Reviewed by the following groups
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
- Evaluated articles (Rapid Reviews Infectious Diseases)
Abstract
Background
Accurate malaria diagnosis is key for patient management, surveillance, and control. Automated microscopy can overcome variation observed among microscopists and is a promising new tool for diagnosis. The Noul miLab integrates smear preparation, staining, imaging, and AI-supported parasite detection in a portable device.
Methods
2201 samples were collected from febrile patients across two sites in Ethiopia, where P. falciparum and P. vivax are frequent, and Ghana, where P. falciparum transmission is intense. Samples were screened by local microscopy at the health center, miLab, and rapid diagnostic test. qPCR and expert microscopy were used as gold standard.
Results
Across the three sites in Ehtiopia and Ghana, miLab reached a sensitivity for P. falciparum diagnosis of 96.3% (335/348) using expert microscopy as gold standard, and of 97.4% (298/306) using qPCR-positive infections at densities >200 parasites/µL as gold standard. Across two sites in Ethiopia, the sensitivity of miLab for P. vivax was 96.8% (399/412) using expert microscopy as gold standard, and 95.9% (419/437) using qPCR-positive infections at densities >200 parasites/µL as gold standard. Specificity compared to qPCR was 98.8% (1057/1070) for P. falciparum and 97.8% (617/631) for P. vivax . The miLab was significantly more sensitive than microscopy conducted at the health center. In Ethiopia, the miLab assigned the correct species to 99.3% (147/148) P. falciparum and 96.5% (304/315) P. vivax mono-infections infections where a species was determined.
Conclusions
The miLab automated microscope shows high sensitivity and specificity for P. falciparum and P. vivax diagnosis.
Article activity feed
-
Strength of evidence
Reviewer(s): C Delahunt (Global Health Labs) | 📗📗📗📗◻️
M Nateghpour (Tehran University of Medical Sciences) | 📒📒📒◻️◻️ -
Charles Delahunt
Review 1: "AI-supported Automated Microscopy for Malaria Diagnosis"
Peer reviewers praised the rigor of the study and its potential to improve access to high quality malaria diagnostics, but urged greater clarity around low parasitemia detection thresholds and cost considerations for real-world deployment.
-
Mehdi Nateghpour
Review 2: "AI-supported Automated Microscopy for Malaria Diagnosis"
Peer reviewers praised the rigor of the study and its potential to improve access to high quality malaria diagnostics, but urged greater clarity around low parasitemia detection thresholds and cost considerations for real-world deployment.
-
