In 2020, when the pandemic known as COVID-19 started, we were all caught off-guard, and that is the reason why we had to put everyone in a lockdown and countries around the world almost stopped for 2-3 months. However, things got back on track slowly and then came the vaccines and in 2023, we can say that the pandemic is well under control except for China which followed a different policy and had its worst outbreak of the virus just a few months back.
Having said that, we are going to focus on the diagnosing aspect of COVID-19 and the fact that we were using nasal swabs to test if someone had the virus or not is considered to be the best. However, there were many companies out there that said that AI could do a better job of detecting if someone caught the virus based on their coughing pattern as well as the sound they make. Without proper data and out of excitement, everyone agreed and said that this AI thing needs to be tried out to see if it really detects COVID-19 better or not.
We finally have an answer, which is not good if you are a fan of this method. According to a recent study made on whether AI detects COVID-19 better or not, it is revealed that AI does a poor job of detecting the virus just based on the cough sounds of a patient. Co-authors of the report said, “The implications are that the AI models used by many apps add little or no value over and above the predictive accuracy offered by user-reported symptoms”
The study’s lead author said that coughs are a poor predictor of COVID-19, which their study has also confirmed. He added, “It’s disappointing that this technology doesn’t work for COVID-19,”. “Finding new ways to quickly and easily diagnose viruses like COVID-19 is important to stop its spread.” With this study, claims from companies such as Fujitsu’s Cough in a Box and others are harder to trust as it is directly related to our health.