An international team of scientists from the University of Washington in Seattle (USA) and the Respiratory Disease Research Center in Nairobi (Kenya) has created an application for mobile devices that allows you to diagnose tuberculosis. The study was published in the scientific journal magazine Science Advances (SciAdv).
A program called TBscreen analyzes human cough sounds using machine learning algorithms.
The cough classification technology was tested by analyzing 33 thousand cases of passive cough and 1.2 thousand cases of specially induced cough in 149 patients with pulmonary tuberculosis and 46 patients with other respiratory diseases.
TBscreen was able to distinguish active TB from non-TB cough with an overall accuracy of approximately 82%. The team found that the app and smartphone microphone were more accurate at predicting which coughs indicated active TB than more expensive devices.
According to the World Health Organization, 10.6 million people worldwide will be infected with tuberculosis in 2022. The disease claimed the lives of more than 1.4 million people.
Previously smart watches Apple Watch rescued The life of an airline passenger.