New Hightech Microscope Using AI Successfully Detects Malaria In Returning Travelers

More than 200 million people are infected with malaria every year and more than half a million die. The World Health Organization recommends diagnosis of specific parasites before starting treatment for diseases caused by Plasmodium parasites. Several diagnostic methods are available, including conventional light microscopy, rapid diagnostic tests, and PCR.

However, the standard for diagnosing malaria is manual light microscopy, where an expert examines blood films under a microscope to confirm the presence of malaria parasites. However, the accuracy of the results depends on the skill of the microscopist, and overworked specialists may be hampered by fatigue as they perform the tests.

Now writing in Frontiers in Malaria , a fully automated system combining artificial intelligence detection software and automated microscopy has assessed malaria with clinically useful accuracy.

"With 88% diagnostic accuracy compared to microscopes, the artificial intelligence system detected almost all malaria parasites, but not as accurately as experts," said researcher Dr. Research has been done. "This level of performance in a clinical setting is a significant achievement for an AI algorithm targeting malaria. It suggests that this system could be a clinically useful tool for diagnosing malaria in appropriate settings."

AI provides accurate diagnosis

Researchers took more than 1,200 blood samples from travelers returning to the UK from malaria-endemic countries. The study effectively tested AI and automated microscope systems in real clinical situations.

They evaluated the samples using both a handheld light microscope and an AI microscope system. By hand, 113 samples were found to be positive for malaria parasites, while the AI ​​system correctly identified 99 samples, corresponding to 88% accuracy.

"AI for medicine often produces encouraging preliminary results on in-house datasets, but real clinical settings are lacking. This study independently assessed whether an AI system would be successful in actual clinical use," he said. Knower, leader. Author of the study.

Automatic vs Manual

The fully automated malaria detection system the researchers are testing includes hardware and software. An automated microscopy platform scans blood spots and malaria detection algorithms to determine the parasite and its current size.

According to scientists, automated malaria testing has several potential benefits. "Even expert microscopists can get tired and make mistakes, especially under heavy workloads," Rees-Chaner explained. "Automated malaria diagnosis using AI could reduce this burden on microscopists and improve patient load." Furthermore, these systems produce reproducible results and are widely applicable, the scientists wrote.

Despite 88% accuracy, the automated system also misidentified 122 samples, potentially leading to patients receiving unnecessary antimalarial drugs. "AI software is not yet as accurate as an expert microscope. This study represents a promising data point, not definitive proof of expertise," said Rees-Channer.

More information: Evaluation of Automated Microscopy Using Machine Learning to Detect Malaria in Returned Travelers to the United Kingdom ( Borderlines in Malaria) (2023). doi: 10.3389/fmala.2023.1148115, www.frontiersin.org/articles/1 … la.2023.1148115/full

Citation : New high-tech microscope using artificial intelligence successfully diagnoses malaria in returning travelers (August 10, 2023) August 10, 2023, https://medicalxpress.com/news/2023-08-high-tech-microscope- Retrieved from ai - malaria.html.

This document is subject to copyright. No part may be reproduced without written permission except for personal research or research fair use. The content is provided for informational purposes only.

The Find Malaria Innovation Platform: Developing Malaria Diagnostics for Wider Country Use