Researchers have developed an AI algorithm that significantly improves the detection of human coughing in electronic devices, reducing false positives and requiring fewer sound samples per second.
A team of engineering and health researchers has developed a new tool that significantly improves the ability of electronic devices to detect when a human patient is coughing. The tool utilises an advanced artificial intelligence (AI) algorithm that allows the AI to express uncertainty, thereby reducing the high rates of false positives encountered in previous cough detection technologies.
This development has potential applications in health monitoring, particularly for people with asthma or COVID-19.
Robust Cough Detection with Out-of-Distribution Detection is published in the IEEE Journal of Biomedical and Health Informatics. According to the researchers, the modified cough detection AI can operate effectively using far fewer sound samples per second than previous technologies, with similar sensitivity and fewer false positives.
This means that the electronic device requires less computing power, making it smaller and more energy-efficient. Additionally, technology will not be recording understandable speech, which addresses privacy concerns.
The modified cough detection AI can operate effectively using far fewer sound samples per second than previous technologies, making it smaller, more energy-efficient, and addressing privacy concerns. The researchers say the approach they’ve taken here could be used to address a range of AI applications in which the AI is likely to encounter unexpected input that it was not trained to understand.
“We’re looking for research partners who can help us explore other health monitoring challenges that this AI modification could help address in a meaningful way,” said Edgar Lobaton, corresponding author of the paper and an associate professor of electrical and computer engineering at North Carolina State University.
The researchers are currently in the process of incorporating the new algorithm into a wearable health monitoring device that can be used in real-world testing. They say that the approach they’ve taken here could be used to address a range of AI applications in which the AI is likely to encounter unexpected input that it was not trained to understand.
The team is looking for research partners who can help explore other health monitoring challenges that this AI modification could help address in a meaningful way.
Ritchelle is a Content Producer for Healthcare Channel, Australia’s premier resource of information for healthcare.
