In an alternative diagnostic method for illness identification, an international team of researchers assessed the precision of an automated microscope combined with AI software to identify malaria parasites in the blood samples of travelers in an actual clinical situation.
Over 200 million people contract malaria annually, and more than 500,000 cases result in fatalities. Before starting treatment for a condition brought on by Plasmodium parasites, the World Health Organization (WHO) advises getting a parasite-based diagnosis.
The researchers collected more than 1,200 blood samples from travelers who had just returned to the UK from countries where malaria is endemic for the new study, which was published in Frontiers in Malaria.
Under ideal circumstances, they put the AI and automated microscopy system to the test in a real clinical scenario.
“At an 88 percent diagnostic accuracy rate relative to microscopists, the AI system identified malaria parasites almost, though not quite, as well as experts,” said Dr. Roxanne Rees-Channer, a researcher at The Hospital for Tropical Diseases at UCLH in the UK.
The ability of AI systems to target malaria to function at this level in a clinical context is a significant accomplishment.
In appropriate conditions, it “indicates that the system can be a clinically useful tool for malaria diagnosis,” Rees-Channer continued.
The researchers used manual light microscopy and AI-microscope equipment to assess samples.
By hand, 113 samples were determined to be positive for the malaria parasite, whereas 99 samples were adequately classified as positive by the AI system, yielding an accuracy rate of 88 percent.
The researchers noted numerous possible advantages of automated malaria diagnosis.
Rees-Channer added that even experienced microscopists “can become fatigued and make mistakes, especially under a heavy workload.”
The stress on microscopists could be lessened by automated malaria diagnosis utilizing AI, enhancing the practical patient load. The scientists also noted that these technologies are broadly deployable and provide reproducible findings.
