September 19, 2022 — Picture this: You believe you may have COVID. Then you speak a few sentences into your phone. An app will then give you reliable results within a matter of minutes.
“You sound sick” is what we humans might tell a friend. Artificial intelligence (or AI) could expand that concept by using your voice to detect COVID infections.
Researchers say that a simple and inexpensive app could be used to screen large crowds at concerts or other large gatherings in low-income countries.
It’s just the latest example in a rising trend exploring voice as a diagnostic tool to detect or predict diseases.
Over the past decade, AI speech analysis has been shown to help detect Parkinson’s disease, posttraumatic stress disorderResearch has been so promising that the National Institutes of Health just launched a new program to study heart disease, dementia, as well as Alzheimer’s. So promising is the research that the National Institutes of Health was just launched. new initiative to develop AI to use voiceA wide range of conditions can be diagnosed. These conditions include respiratory diseases such as pneumoniaand COPD for laryngeal cancer, and even StrokeALS, and psychiatric disorders, such as schizophrenia and depression. Software can detect nuances that the human ear can’t, researchers say.
This approach to COVID detection has been employed in at least half of the studies. Researchers at Maastricht University, Netherlands, reported that their AI model was accurate 89%, as compared to an average accuracy of 56% for various other lateral flow tests. In addition to detecting infection, people who are not showing symptoms were also more accurately detected by the voice test.
One thing to note: The false positive rate for lateral flow tests is less than 1%, while it shows 17% for voice tests. Still, since the test is “virtually free,” it would still be practical to just have those who test positive take further tests, Wafaa aljbawi, researcher, who presented the preliminary findings at the European Respiratory Society’s International Congress in Barcelona, Spain.
“I am personally excited for the possible medical implications,” says Visara Urovi, PhD, a researcher on the project and an associate professor at the Institute of Data Science at Maastricht University. “If we better understand how voice changes with different conditions, we could potentially know when we are about to get sick or when to seek more tests and/or treatment.”
The development of the AI
A COVID infection can change your voice. It affects the respiratory tract, “resulting in a lack of speech energy and a loss of voice due to shortness of breath and upper airway congestion,” says the preprint paper, which hasn’t been peer reviewed yet. A COVID patient’s typical dry cough also causes changes in the vocal cords. And previous research found that lung and larynx dysfunction from COVID changes a voice’s acoustic characteristics.
The dataset is large is part of what makes this research unique. Researchers used a crowd-sourced database at the University of Cambridge, which contained 893 audio samples of 4,352 people. Of these, 308 were positive for COVID.
You can contribute to this database – it’s all anonymous — via Cambridge’s Appreciation for COVID-19 SoundsThis question asks you to You can coughRepeat the process three times. Breathe deeply through the mouthRead a sentence three times and repeat the process three to five more times.
For their study, Maastricht University researchers “only focused on the spoken sentences,” explains Urovi. The “signal parameters” of the audio “provide some information on the energy of speech,” she says. “It is those numbers that are used in the algorithm to make a decision.”
Audiophiles might find it fascinating that the researchers used Mel spectrogram analysis (mel spectrogram analysis) to identify characteristics in the sound wave (or the timbre). Artificial intelligence enthusiasts will appreciate the fact that the study revealed that long-term memories (LSTM), was the most effective type of AI model. It’s based on neural networks that mimic the Brain of the human beingIt is particularly adept at modeling signals that have been collected over time.
For laypeople, it’s enough to know that advancements in the field may lead to “reliable, efficient, affordable, convenient, and simple-to-use” technologies for detection and prediction of disease, the paper said.
What’s Next?
Urovi says that validation is essential for transforming this research into an app. Such “external validation” — testing how the model works with another dataset of sounds — can be a slow process.
“A validation phase can take years before the app can be made available to the broader public,” Urovi says.
Urovi stresses that even with the large Cambridge dataset, “it is hard to predict how well this model might work in the general population.” If speech testing is shown to work better than a rapid antigen test, “people might prefer the cheap non-invasive option.”
According to the paper, “But further research is required to explore which voice features can be most helpful in picking COVID patients and make sure models are able to tell the difference between COVIDs and other respiratory conditions,”.
Pre-concert app testing is also possible in the near future. That’ll depend on cost-benefit analyses and many other considerations, Urovi says.
Nevertheless, “It may still bring benefits if the test is used in support or in addition to other well-established screening tools such as a PCR test.”
