What if you just had to cough on a smartphone for it to tell you what disease you have? The American start-up Hyfe believes in it and is currently preparing an application in this direction.
According to the very serious Wall Street Journal, the start-up is based on the construction of an algorithm capable of distinguishing the sound and the frequency of the cough and affirms that each respiratory disease which forces us to cough offers “its own signature”. “This is called epidemiological acoustics,” explains Peter Small, tuberculosis expert and medical director of the start-up, at american newspaper.
And the choice to analyze the different coughs is not trivial, since these are generally among the first symptoms of a disease, whether it is a simple chronic cough, pneumonia, bronchitis, seasonal flu, asthma, tuberculosis or the now famous covid-19.
No self-medication
Still in the experimental phase, the work carried out by Hyfe should ultimately lead to the creation of two mobile applications, one intended for the general public and the other for researchers. But the goal is not to allow everyone to self-diagnose and then engage in self-medication. Act which remains very strongly discouraged by health professionals.
Their creators explain that the uses could be multiple, especially in countries where tuberculosis is still very present. After a first test on an equipped smartphone, “people who would test positive on the application would then undergo a laboratory test to see if they really have tuberculosis,” says Grant Theron, professor at Stellenbosch University in Cape Town ( South Africa) in the Wall Street Journal. Clinical trials are currently underway with some 800 patients, in various universities and research centers around the world, including at the University of Navarre and the Zizur Health Center in Spain, or the University Hospital Center of Montreal in Canada.
This technology based on listening to a patient’s cough is not new. Last year, the famous MIT published the first results of its work for an application to analyze coughs to say whether or not a person has Covid-19, according to a study published in the journal IEEE Open Journal of Engineering in Medicine and Biology. Of the 200,000 cough sound samples, the algorithm had accurately identified 98.5% of people confirmed to have Covid-19 and correctly screened 100% of asymptomatic carriers.
