Malaria remains a leading health threat in sub-Saharan Africa, causing over 600,000 deaths annually. Ghana, ranked among the top 15 in malaria burden in Africa, relies heavily on microscopy for diagnosis due to its affordability and established use. However, limitations in sensitivity, turnaround time, and availability of skilled personnel despite ongoing national training efforts stress the need for improved diagnostics. This study evaluated miLab™, an AI-assisted automated malaria detection platform, using nested PCR (nPCR) as the reference standard. We conducted a hospital-based cross-sectional study from August 2024 to June 2025 in three malaria-endemic communities in Kumasi, Ashanti region and enrolled 300 suspected malaria patients (168 females, 132 males; aged 1-87 years, median 24). Blood samples were analyzed independently by miLab™, two independent mid-level microscopists, and nPCR. Discrepancies between the two microscopists were resolved by a WHO expert microscopist (adjudicated microscopy) to establish a microscopy reference standard. Diagnostic accuracy, correlation and measurement agreement were determined using GraphPad Prism Version 8. Parasite densities estimated by miLab™ ranged from 0.95 to 5.34 log parasites/µL (median 3.52, IQR 2.93-3.91). For mid-level microscopist 1, densities ranged from 2.02 to 5.43, (median 4.02, IQR: 3.22-4.50), while Mid-level microscopist 2 measured densities between 1.98 and 5.22, (median 3.78, IQR: 3.15-4.23). When compared to nPCR, miLab™ demonstrated a sensitivity of 94.23%, specificity of 98.98%, and accuracy of 97.33%, while adjudicated microscopy showed a sensitivity of 85.58%, specificity of 97.96%, and accuracy of 93.67%. These findings indicate that miLab™ has better performance compared to microscopy for detection of malaria parasites. miLab™ could offer a reliable, rapid diagnostic alternative suitable for malaria-endemic, resource-limited settings where timely and accurate diagnosis is critical for effective case management and control.
Keywords: AI-assisted device; Ashanti region; Automated microscopy; Ghana; Malaria diagnosis; MiLab™; nPCR.
© 2026. The Author(s).