Artificial Intelligence Continues to Assist in Cough Detection

A new artificial intelligence algorithm is making it possible to detect coughs in more real-world settings, according to a paper recently published in the IEEE Journal of Biomedical and Health Informatics.

The paper “Robust Cough Detection with Out-of-Distribution Detection” said the use of AI to detect coughing is not unprecedented. Despite its previous applications, researchers noted that there are limitations that generally exist, such as the training of algorithms on impractical data.

 “When AI is being trained to identify the sound of coughing, this is usually done with ‘clean’ data – there is not a lot of background noise or confusing sounds,” Edgar Lobaton, corresponding author of a paper on the work and an associate professor of electrical and computer engineering at North Carolina State University, said in a press release.

“But the real world is full of background noise and confusing sounds. So previous cough detection technologies often struggled with ‘false positives’ – they would say that someone was coughing even if nobody was coughing,” Lobaton continued.

While considering these facts, Lobaton and co-authors created a new tool that allows the AI resource to comprehend and note foreign sounds.

“We’ve developed an algorithm that helps us address this problem by allowing an AI to express uncertainty. Rather than having to decide ‘Yes, that was a cough’ or ‘No, that wasn’t a cough,’ the AI can also report that it has detected a sound it’s not familiar with. In other words, the AI is given a third option: ‘I don’t know what that was,’” stated Lobaton.

Although this is not the first occurrence of this type of device, the false positives associated with previous devices often contain reports of coughs related to sounds that are not coughs.

However, the new algorithm took part in a test within computational models, leading researchers to conclude that it can operate using fewer sound samples per second.

Furthermore, researchers are aiming to add the tool to a wearable health monitoring device. They noted that this system has the potential for broad application, as this type of technology could be highly effective in treating various conditions.

“For example, there is interest in using wearable health monitoring devices that would detect coughs in people who have asthma, which could trigger a notification about increased risk of an asthma attack,” said Lobaton. “There is also interest in using cough detection for COVID monitoring, and so on.”

The use of AI to support the detection of symptoms for various conditions is a common and growing practice.

An example of this took place in March when researchers from Stanford University and other institutions created an AI-based app that could capture pictures of skin lesions . The app provides researchers with insight as to whether they were related to mpox.

After capturing the photos, the system would provide patients with a risk score and recommendations for testing or vaccination.