Automated AI-supported microscopy and portable qPCR for the diagnosis of Plasmodium falciparum and Plasmodium vivax malaria

C. Koepfli
University of Notre Dame,
United States

Summary:

Sensitive and accurate malaria diagnosis is required for case management to accelerate control efforts. Diagnosis is particularly challenging where multiple Plasmodium species are endemic. The Noul miLab is a fully automated portable digital microscope that prepares a blood film from a droplet of blood, followed by staining and detection of parasites by AI. Infected red blood cells are displayed on the screen of the instrument. Time-to-result is approximately 20 minutes, with less than two minutes hands-on time. We evaluated the miLab among 659 suspected malaria patients in Gondar, Ethiopia, where P. falciparum and P. vivax are endemic, where P. falciparum transmission is intense. We developed a portable qPCR platform for on site parasite screening and quantification as gold standard. Across both countries combined, the sensitivity of the miLab for P. falciparum was 94.3% at densities >200 parasites/µL by qPCR, and 83% at densities >20 parasites/µL, and specificity was 94.0%. For P. vivax diagnosis in Ethiopia, the sensitivity of the miLab was 97.0% at densities >200 parasites/µL (RDT: 76.8%, microscopy: 67.0%), 93.9% at densities >20 parasites/µL, and specificity was 97.6%. In conclusion, the miLab was more sensitive than microscopy and thus is a valuable addition to the toolkit for malaria diagnosis, particularly for areas with high frequencies of hrp2/3 deletions.