Tools & Strategies News

Artificial Intelligence Enhances Pediatric Tuberculosis Diagnosis Process

Research shows that a test combining nanotechnology and artificial intelligence can improve the tuberculosis diagnosis process among pediatric patients, detecting 89 percent of cases.

AI for disease tracking.

Source: Getty Images

By Mark Melchionna

- Developed by researchers at Tulane University and described in a study published in Nature Biomedical Engineering, a new blood testing system displayed the ability to enhance the pediatric tuberculosis diagnosis process through artificial intelligence (AI) and nanotechnology.

Tuberculosis (TB) is caused by a bacterium that usually attacks the lungs but can also attack any part of the body, like the kidney and brain. According to the press release, recent data indicates that the only infectious disease that currently surpasses TB in terms of prevalence is COVID-19. There were 7,860 TB cases in the US in 2021, a rate of 2.4 per 100,000 persons.

Each year, about a quarter of a million children younger than 15 die from TB, 80 percent of whom are younger than 5, the press release states. The majority of those cases (96 percent) did not receive a diagnosis. Further, TB can be deadly for young children with HIV.

Tony Hu, PhD, the Weatherhead Presidential Chair in Biotechnology Innovation at Tulane University, noted that technology-based TB diagnosis and treatment methods were worth examining due to the high number of limited-access communities battling TB.

“TB is a disease found primarily in resource-limited areas, the ideal is to create a smartphone-based method that could be used at the point-of-care in these settings,” said Hu, in a press release.

The AI-and nanotechnology-based blood test developed by Tulane Researchers provides a view into lipoarabinomannan (LAM) and LprG, which make up the bacteria that cause TB. The molecule and associated protein are found in membrane-bound sacs, called extracellular vesicles, that are shed by the immune cells of TB patients.

Hu and his research team located these vesicles by coating nanoparticles with antibodies that bind LAM and LprG. 

To ensure the test had a high degree of sensitivity, they also created an AI algorithm to eliminate unneeded information, or 'background noise,' caused by other substances on the nanoparticle, according to the press release. Overall, the test accurately detected TB in 89 percent of children who were already diagnosed with the disease. It also identified TB in 74 percent of children who had received standard test results that did not indicate TB.

Recently, the use of AI has grown rapidly, with providers and organizations finding it useful in various efforts.

In August, researchers from the University of Florida created an AI algorithm to track COVID-19 variants. They engaged in this effort to mitigate the challenges of keeping up with COVID-19 and the problems that new variants can cause. Using data on the genetic sequences of COVID-19, researchers trained an AI algorithm to flag potential new variant that attacks cells.

Another study from April found that an AI-based system helped clinicians determine the treatment that bladder cancer patients may need after chemotherapy. After using three measurement methods for determining patient responses to chemotherapy, researchers used the AI system to determine the accuracy and abilities of each measure.