This Technology Can Allegedly Identify COVID-19 Through Your Voice Characteristics: How Does This Work

The COVID-19 pandemic has led to many discoveries including vaccines and special tests, which help medical experts alleviate the problem. Previously, there was an AI tool used to predict the potential animal host of the next coronavirus.

Indeed, creating a helpful tool does not stop, as recently, an AI health tech company called Vocalis has arrived with a possible solution to screen people.

Vocalis has apparently created an AI technology that could allegedly predict if someone has the virus through their voice.

COVID-19 Screening Using Voice Characteristics

This Technology Can Identify COVID-19 Through Your Voice Characteristics: How Does This Work

(Photo : Anna Shvets from Pexels)

It is a common scenario for people to go to the nearest hospital or screening center if they want to know whether they have COVID-19 or not. However, at the moment, going outside the house might be risky for everyone, as COVID-19 could spread in the air, HealthLine reported.

It is possible to get some COVID-19 detection kit at home, but not everyone has access to it.

Since the pandemic erupted last 2020, experts and other technologists created a solution to hasten the detection of the virus among humans.

Vocalis, a known company that incorporates health to AI, has developed a technology that makes use of a person’s voice to allegedly predict whether they have been infected. The process is pretty simple, and it starts with the company collecting the data from the subject’s voice.

The technology would then identify the characteristics of COVID-19 by analyzing the data gathered from the voice.

Specifically, 512 features are extracted from a vocal sample, and they are interpreted by the spectrogram as images.

In conducting clinical tests, Vocalis first needs to know the type of voice recordings that would be collected for the identification of the disease. It highlights the information on the audio, which will be carried out in the trials.

To understand how the method works, a person who has COVID-19 symptoms like shortness of breath, fever, and headache undergoes a detection test using AI technology to identify the symptom’s underlying characteristics.

Read Also:  Pfizer vs Moderna: Comparing Two COVID-19 Vaccines from Side Effects to Storage Requirements

Vocal Biomarkers Are Used to Enable COVID-19 Detection Technology

This Technology Can Identify COVID-19 Through Your Voice Characteristics: How Does This Work

(Photo : Screenshot from YouTube/Aseem Chopra)

According to The Next Web, the company ensures that the AI being used is language agnostic. In algorithm training, the person will be instructed to count from 50 to 70.

Vocalis’ co-founder, Dr. Shady Hasan stated that the technology respects people’s privacy by keeping their identities confidential. This also guarantees that no information will be shared.

Moreover, the company also aims to be consistent in its data gathering. To achieve consistency in its data set, Vocalis makes use of vocal biomarkers that help the experts in detecting a particular disease of a patient. 

It is said that more than 275,000 various speakers of different languages are contained in the pre-training dataset. Vocalis want the solution to be accessible to everyone around the world.

More importantly, the accuracy of the algorithm technology is said to be 81.2%, and that includes its ability to identify COVID-19 traces from an asymptomatic patient.

Related Article: This AI Can Forecast the New Animal Host of the Next Coronavirus

admin

Leave a Reply

Your email address will not be published. Required fields are marked *

Next Post

Casio Launches Latest G-Shock Equipped with Wear OS, Sensors, and More

Fri Apr 2 , 2021
Casio G-Shock Wear OS added   ( Youtube/ NukeForTech ) Rumors have been going around for some time now that there will be a rugged Apple Watch, but Casio has beat Apple to it. Casio has unveiled its first G-Shock rugged smartwatch using Google’s Wear OS platform. Casio G-Shock Wear OS According […]