Vaishnavi VL

5th Pharm D

VOCAL BIOMARKER TECHNOLOGY : Voice analysis to detect COVID 19

25 July 2021

COVID 19 tests are an essential element in diagnosing infectious diseases and assessing whether someone is already infected. However, the lack of available tests and the high level of accuracy required from them have dampened expectations for large-scale use against the corona virus. However, advanced techniques have been developed to detect COVID 19 by using the patient's sound waves, is known as Vocal biomarkers.

 

Vocal Biomarker

A biomarker is an object that is objectively measured and evaluated and represents a biological or pathological process or pharmacological response to a therapeutic intervention, as well as can be used as a surrogate marker at a clinical endpoint. But a vocal biomarker is a signature, feature, or combination of features from an audio signal in the background of the sound; it is also used to monitor patients, diagnose or grade them, and to determine the severity or stages of a disease associated with a clinical outcome.

 

Types of Voice Recordings

There is no standard voice protocol for identifying vocal biomarkers, but one can classify the sounds emitted by the human mouth and analyse them into 3 main categories for diagnosis: verbal (isolated words, short sentence repetition, reading speech). Sustained vowel phonations are a type of recording, in which the voice of a vowel is requested to be retained for as long as possible. Sustained vowel phonation carry information to assess dysphonia, and enable the patient's voice to be measured without obvious effects, without being affected by speech rate, stress, or vocal cords.

 

Data Collection Techniques

Different data collection methods have been developed. They can be divided into 4 main categories:

 

1. Studio-based recording involves recording speech into a controlled environment, which minimizes unnecessary phonology and avoids proximity effects. This often leads to an exaggeration of low-frequency sounds due to the proximity of the sound source from the microphone. Generally, the recommended distance is 15 to 30 cm.

 

2. Telephone-based recording requires data collection from various speakers and handsets, where there are a number of disadvantages, such as handset noise, lack of control over the speaker environment, and bandwidth limitations.

 

3. Web-based recording is a very popular technology for large-scale data collection campaigns and relies on Internet access to make it easily accessible.

 

4. Smartphone based recording provides broadband quality with smartphone devices that are widely available and affordable. Smartphone / web based recording has similar disadvantages to telephone based recording in addition to the bandwidth limit.

 

The Science behind the Smartphone Based COVID-19 Voice Test

 

Whenever a human lung is affected by a disease, their voice is the first to be affected. Although the change is not visible to the human ear due to oral limitations, a computer can easily operate on various parameters of the voice sample and decode any changes in it. Because the corona virus affects the lungs, the science behind the COVID-19 voice test, based on a smartphone, suggests that the COVID-19 test will also evaluate effective results through voice samples.

 

How does the COVID-19 test work on a smartphone?

 

The patient can record their voice sample by downloading and installing the Voice Analysis application on a smartphone for the COVID-19 test. The application, with the help of Internet servers, matches the patient's voice sample to the voices of millions of people tested negative for COVID - 19, and samples are already loaded into the application. The result of this test will appear in just 30 seconds.

To achieve this incredible achievement, the voice analysis application enlists the help of Artificial Intelligence (AI). According to the latest developments in data science, a human voice contains 6300 parameters. Based on this data, the AI will work and examine the voice sample of the associated COVID-19 patient.

Naturally, voice recognition will only be a preliminary test before more accurate tests are performed. However, this will significantly speed up the detection process.COVID 19 tests are an essential element in diagnosing infectious diseases and assessing whether someone is already infected. However, the lack of available tests and the high level of accuracy required from them have dampened expectations for large-scale use against the corona virus. However, advanced techniques have been developed to detect COVID 19 by using the patient's sound waves, is known as Vocal biomarkers.

 

20
It will be great, if you share your view on above write-up. 

Your content has been submitted. Please refresh page to see comment