This new COVID test has a turnaround time of about a minute and can enable remote, virtual testing. They could be used at huge gatherings to filter the population quickly.
A mobile phone app with the help of artificial intelligence (AI) can help to detect COVID-19 infection in people’s voices, says a new study presented at the European Respiratory Society International Congress in Barcelona, Spain.
This study’s AI model outperforms lateral flow/rapid antigen testing in terms of accuracy, plus it’s inexpensive, fast, and simple to use, making it ideal for usage in low-income nations where PCR tests are extremely costly or otherwise difficult to deliver.
At the congress, Ms. Wafaa Aljbawi of Maastricht University’s Institute of Data Science reported that the AI model was accurate 89% of the time, whereas the accuracy of lateral flow testing varied greatly from brand to brand. Additionally, lateral flow tests had a significantly lower ability to detect COVID infection in those with no symptoms.
“These promising results suggest that simple voice recordings and fine-tuned AI algorithms can potentially achieve high precision in determining which patients have COVID-19 infection,” she added.
Additionally, they support virtual, remote testing and have a turnaround time of under a minute. They could be utilized, for instance, at the entrances to major gatherings to enable quick populace screening.
When a person has COVID-19, it usually affects their upper respiratory tract and vocal cords, which can change their voice. Ms Aljbawi and her supervisors, Dr Sami Simons, pulmonologist at Maastricht University Medical Centre, and Dr Visara Urovi, also from the Institute of Data Science, wanted to see if AI could be used to analyze speech in order to diagnose COVID-19.
They used information from the crowdsourced COVID-19 Sounds App from the University of Cambridge, which includes 893 audio samples from 4,352 healthy and unhealthy subjects, of which 308 received COVID-19 positive test results. After downloading the app, participants fill out a short survey about their demographics, health, and smoking habits before being asked to capture their breathing sounds. Coughing three times, taking three to five deep breaths through their mouths, and reading a brief sentence on the screen three times are a few of these.
The researchers employed a method for analyzing voice known as Mel-spectrogram analysis, which distinguishes several voice characteristics like loudness, power, and fluctuation across time.
“In this way, we can decompose the many properties of the participants’ voices,” Ms Aljbawi said. “In order to distinguish the voice of COVID-19 patients from those who did not have the disease, we built different artificial intelligence models and evaluated which one worked best at classifying the COVID-19 cases.”
After comparing the results of various models, they concluded that Long-Short Term Memory (LSTM) was the most effective. Neural networks, which replicate the way the human brain functions and recognize the underlying correlations in data, are the foundation of LSTM. It can store data in its memory and works with sequences. This makes it good for modeling signals that are collected over time, like the voice.
The overall accuracy was 89%, the sensitivity was 89%, and the specificity was 83%, meaning that it properly detected 89% of positive instances and 83% of negative cases.
Compared to cutting-edge procedures like the lateral flow test, the results “show a significant improvement in the accuracy of diagnosing COVID-19,” said Ms. Aljbawi. Although the lateral flow test’s specificity rate is greater at 99.5%, its sensitivity is only 56%.
Although the lateral flow test’s specificity rate is greater at 99.5%, its sensitivity is only 56%.
“This is important as it signifies that the lateral flow test is misclassifying infected people as COVID-19 negative more often than our test. In other words, with the AI LSTM model, we could miss 11 out 100 cases who would go on to spread the infection, while the lateral flow test would miss 44 out of 100 cases.
“The high specificity of the lateral flow test means that only one in 100 people would be wrongly told they were COVID-19 positive when, in fact, they were not infected, while the LSTM test would wrongly diagnose 17 in 100 non-infected people as positive. However, since this test is virtually free, it is possible to invite people for PCR tests if the LSTM tests show they are positive.”
According to the study authors, further research with larger samples is required to confirm their findings. 53,449 audio samples from 36,116 individuals have been collected since the experiment began and can be used to enhance and validate the model’s accuracy. They are also conducting additional research to determine which aspects of the speech are influencing the AI model.
Image Credit: Getty
You were reading: Your Voice Could Reveal If You Are Exposed To COVID Within A Minute – Here’s How To Check