Effect of Artificial Intelligence-Based Technology in Malaria Diagnosis

Abstract

BACKGROUNDMalaria diagnosis remains a key issue in remote area. The gold standard microscopy has limitations of its own and the commonly used rapid diagnostic tests have variable results. An artificial intelligence based digital cytometry platform was assessed for its comparison with microscopy and rapid diagnostics test in terms of accuracy and feasibility. METHODS100 patients with complaints of fever with chills were included in the study. They were screened for malaria infection by microscopy of thin peripheral smear, antigen based rapid diagnostic test and digital cytometry. RESULTSAt higher malaria parasite concentration, the results between microscopy and digital cytometry were matching exactly. But at lower parasite concentration microscopy had lower sensitivity than digital cytometry. When compared to rapid diagnostic tests the digital cytometry had advantage of better species identification and better sensitivity. CONCLUSIONSIn our study, digital cytometry has greater sensitivity at lower parasite concentration as compared to rapid diagnostic test and microscopy. It was more accurate in species identification as compared to rapid diagnostic test.