People have learned to coexist with COVID-19. While we are almost back to normal, coronavirus cases are still being recorded and how. People no longer need to wait in lines for COVID tests due to technological advancements. There are also a number of home test kits available. What if we said a smartphone app could perform a COVID test? Isn’t that surprising? According to a study. That a smartphone app can accurately detect COVID-19 infection.
AI to detect COVID-19 infection in people’s voices
Researchers at Maastricht University’s Institute of Data Science in the Netherlands have created a smartphone app that uses AI (artificial intelligence) to detect COVID-19 infection in people’s voices. That’s right, there’s no need for a nasal specimen.
The app, according to the researchers, is more accurate than several antigen tests and is inexpensive, quick, and simple to use. This means that it can now be used in low-income countries. The researchers also stated that this software can be used in countries where PCR tests are relatively expensive or distribution by the government is difficult.
Wafaa Aljbawi, a researcher at the Institute of Data Science, Maastricht University, The Netherlands said, “The 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.” “Moreover, they enable remote, virtual testing and have a turnaround time of less than a minute. They could be used, for example, at the entry points for large gatherings, enabling rapid screening of the population,” Aljbawi added during her presentation at the European Respiratory Society International Congress in Barcelona, Spain.
Used method for analysing voice known as Mel-spectrogram analysis
During her time working on the project, Aljbawi and her colleagues first investigated whether it was possible to use AI to analyse voices in order to detect COVID-19 infection. The team is said to have used data from the crowdsourced COVID-19 Sounds App from the University of Cambridge, which includes 893 audio samples from 4,352 healthy and unhealthy subjects. 308 of them tested positive for COVID-19. The researcher then used a method for analysing voice known as Mel-spectrogram analysis, which identifies several voice characteristics such as loudness, power, and fluctuation over time.
“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,” Aljbawi explained. The research team discovered that the Long-Short Term Memory (LSTM) performed better than the others. The model is based on Neural networks, which replicate the way the human brain functions and recognizes the underlying patterns in data.
The team revealed that the overall accuracy of the app was recorded at 89 per cent. The app was around 89 per cent accurate in identifying positive COVID-19 cases and 83 percent accurate in identifying the Covid-19 negative cases. “These results show a significant improvement in the accuracy of diagnosing COVID-19 compared to state-of-the-art tests such as the lateral flow test,” said Aljbawi.
Notably, the team is still conducting the tests and their results still need more tests and findings based on a larger populace.
