London, Sep 5 (PTI) Scientists have developed a new smartphone app that can accurately detect COVID-19 infection in people’s voices using artificial intelligence (AI).
The AI model used in the research is more accurate than rapid antigen tests or lateral flow tests and is cheap, quick and easy to use, the researchers said.
The method can be used in low-income countries where PCR tests are expensive and difficult to distribute, they said.
The discovery was presented Monday at the International Congress of the European Respiratory Society in Barcelona, Spain.
According to the researchers, the AI model is accurate 89% of the time, while the accuracy of lateral flow tests varies widely by brand.
Additionally, lateral flow tests are considerably less accurate at detecting COVID-19 infection in people who have no symptoms, they said.
“These promising results suggest that simple voice recordings and fine-tuned AI algorithms can potentially achieve high accuracy in determining which patients are infected with COVID-19,” Maastricht University researcher Wafaa Aljbawi told reporters. Netherlands.
“Such tests can be provided free of charge and are simple to interpret. Moreover, they allow remote virtual testing and have a turnaround time of less than a minute,” Aljwabi said.
The new test could be used, for example, at entry points to large gatherings, allowing rapid screening of the population, the researchers said.
COVID-19 infection typically affects the upper respiratory tract and vocal cords, causing changes in a person’s voice.
Aljbawi and his supervisors used data from the University of Cambridge’s crowd-sourced COVID-19 Sounds app which contains 893 audio samples from 4,352 healthy and unhealthy participants, 308 of whom had tested positive for COVID-19.
The application is installed on the user’s phone. Participants report some basic information about demographics, medical history, and smoking status, and then are asked to record some breath sounds.
These include coughing three times, taking a deep breath through your mouth three to five times, and reading a short sentence on the screen three times.
The researchers used a voice analysis technique called Mel spectrogram analysis, which identifies different voice characteristics such as loudness, loudness and variation over time.
“That way we can decompose the many properties of participants’ voices,” 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 in classifying COVID-19 cases,” he said. -she adds.
They found that a model called long-term memory (LSTM) outperformed the other models. LSTM is based on neural networks, which mimic the functioning of the human brain and recognize underlying relationships in data.
Its overall accuracy was 89%, its ability to correctly detect positive cases or “sensitivity” was 89% and its ability to correctly identify negative cases, or “specificity” was 83%, the researchers found.
In another study, Henry Glyde, a PhD student at the University of Bristol, showed that AI could be harnessed through an app called myCOPD to predict when patients with chronic obstructive pulmonary disease (COPD) might experience a flare-up. of their illness. COPD exacerbations can be very serious and are associated with an increased risk of hospitalization.
Symptoms include shortness of breath, coughing, and the production of more mucus. PTI SAR SAR
(Disclaimer: This story is auto-generated from a syndicated feed; only image and title may have been reworked by www.republicworld.com)