Evaluation of an automated microscope using machine learning for the detection of malaria in travelers returned to the UK

• DOI: 10.3389/fmala.2023.1148115
• Publication Date: 2023-08-10T08:22:28Z
• Authors: Roxanne R. Rees-C...; Christine M. Bachman; Lynn Grignard; Michelle L. Gatton; Stephen Burkot; Matthew P. Horning; Charles B. Delahunt; Liming Hu; Courosh Mehanian; Clay M. Thompson; Katherine Woods; Paul Lansdell; Sonal Shah; Peter L. Chiodini
• Keywords: Giemsa stain; Diagnosis of malaria; Blood smear

Introduction Light microscopy remains a standard method for detection of malaria parasites in clinical cases but training to expert level requires considerable time. Moreover, excessive workflow causes fatigue and can impact performance. An automated tool could aid clinics with limited access highly skilled microscopists, where case numbers are excessive, or multi-site studies consistency is essential. The EasyScan GO an scanning microscope combined machine learning software designed detect field-prepared Giemsa-stained blood films. This study evaluates the ability detect, quantify identify species parasite present films compared light microscopy. Methods Travelers returning UK testing positive were screened eligibility enrolled. Blood samples from enrolled participants used make smears assessed by determine density species. also PCR confirm resolve discrepancy between manual GO. Results When microscopy, exhibited sensitivity 88% (95% CI: 80-93%) specificity 89% 87-91%). Of 99 labelled both, identified 87 as Plasmodium falciparum ( Pf ) 12 non- . correctly reported 86 11 samples. However, it failed distinguish species, reporting all P. vivax calculated densities within +/-25% 33% 200 2000 p/µL, falling short WHO 1 (expert) competency (50% should be true parasitemia). Discussion shows that proficient detecting relative accurately Performance at low densities, distinguishing accurate quantitation parasitemia require further development evaluation.