A team of specialists from several institutions is employing artificial intelligence (AI) to diagnose parasitic infections in patients by scanning stool samples.

Existing methods for locating parasitic worm eggs in stool samples are typically conducted by trained lab technicians. Yet, those living in remote regions who have neither access nor funds to pay for such tests, go untested and subsequently, untreaded.

Source: PLOS Neglected Tropical Diseases. DOI: 10.1371/journal.pntd.0012041Source: PLOS Neglected Tropical Diseases. DOI: 10.1371/journal.pntd.0012041

As such, the researchers trained and tested an AI application on roughly 1,300 stool samples from children in Kenya. Following local processing, the samples were digitally scanned via microscope cameras. Those scans were then uploaded to the cloud, where they were available for use by the AI app.

Focusing exclusively on three types of parasitic infections — hookworms, roundworms and whipworms —the researchers diagnosed infections through the discovery of worm eggs in processed fecal samples.

The team reported that the AI app successfully detected the presence of eggs in 76% to 96% of infections, with results varying according to the types of eggs involved. Further, the app only produced false identifications roughly 1% to 2% of the time.

Importantly, the team explained that the AI analysis of a sample took anywhere from five minutes to half an hour, depending upon upload speed.

The article detailing the app,Diagnosis of soil-transmitted helminth infections with digital mobile microscopy and artificial intelligence in a resource-limited setting,” appears in the journal PLOS Neglected Tropical Diseases.

To contact the author of this article, email mdonlon@globalspec.com