Diagnostic Accuracy of Utilizing Artificial Intelligence for Malaria Diagnostic: A Systematic Review and...

Malaria remains a major public health concern around the world. Microscopic blood smear examination continues to be the gold standard for diagnosis; however, it requires high technical skills and expertise, limiting diagnostic accuracy in resource-poor settings. Artificial intelligence (AI) has emerged as a promising tool to support malaria detection. This systematic review provides an overview of the diagnostic performance of AI-based systems for malaria diagnosis in a clinical setting.

This study followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines and involved articles within the last 10 years that were collected from PubMed, ScienceDirect, Cochrane, EBSCO, and Wiley Online Library. Original articles that reported AI diagnostic accuracy with external validation were involved. The quality of each study was evaluated using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2).

Ten studies with 6754 patients were analyzed. Pooled results of sensitivity [87.7% (95% CI: 78.2-93.4)] and specificity [91.4% (95% CI: 77.3-97.1)] revealed how much the AI agrees with each method when that method is used as a gold standard. Additionally, AI achieved a sensitivity of 87.7% and a specificity of 91.4% compared to microscopy examination and a sensitivity of 90.7% and a specificity of 88.3% compared to polymerase chain reaction (PCR).

AI-based systems improve malaria diagnosis by providing high accuracy, automation, and lower costs. Showing performance comparable to reference methods such as microscopy and PCR, AI is a promising complementary tool for malaria control.