2024-03-02 15:52:28
In Kenya, the Medical Research Institute is developing a mobile application that uses artificial intelligence to diagnose tuberculosis and other respiratory diseases.
The researchers’ goal is to create a mobile application that can accurately recognize cough linked to tuberculosis and other serious illnesses. Nduba Fidelis, the lead researcher, explains that “this software sort of uses artificial intelligence and tries to analyze coughs, which we call cough spectral grams, and then regress them. So it’s a mathematical way of modeling the image of coughing to determine if there’s a difference between someone who has TB and someone who doesn’t when they cough.”
The institute’s app collects coughs using three microphones, the results are then sent to the University of Washington, which runs them through existing software called ResNet 18. Nduba believes that if the software can be tested accurately, it can reduce the time it takes to get a diagnosis and treatment, which will help stop the spread of TB. Indeed, “between the time a person shows symptoms of tuberculosis and the time a doctor determines that they have the disease and need treatment, there is an average of 3 to 2 month, or even a year. » This application which detects tuberculosis makes it possible to “reduce transmission within the community, because a large part of tuberculosis is due to transmission”, he explains. The trial was funded by the National Institutes of Health, but it has not yet received regulatory approval.
Par WritingEditorial Committee
1709397971
#mobile #application #detect #tuberculosis
