BMC to start screening of suspected Covid-19 patients with voice samples, AI from Wednesday

Brihanmumbai Municipal Corporation (BMC) will start screening suspected Covid-19 patients through artificial intelligence (AI), using their voice samples from Wednesday. On the occasion of Ganesh Chaturthi, the civic body signed the memorandum of understanding (MOU) with an Israel-based startup company, Vocalis Health, for the first-of-its-kind initiative in Mumbai.

HT Image

Vocalis Health is collecting voice samples of people across the globe as an initiative to develop a diagnostic tool to identify Covid-19 infected people. Even though the concept is new, several countries like the US and Israel are using it. People can submit their voice samples for analysis on the company’s official website.

From Wednesday, the pan-worldwide study will be initiated in Mumbai at the NESCO jumbo Covid centre, Goregaon. Over 2,000 individuals suspected of having Covid-19 or high-risk contacts of infected patients will be tested with their voice samples.

“We signed the MOU on Saturday and the study will start from Wednesday under the supervision of Dr RN Cooper Municipal General Hospital,” said Suresh Kakani, additional commissioner, BMC. “It will take two to three months to conclude the study,” he said.

The gold standard test for diagnosis of the Sars-Cov-2 virus, which causes Covid-19, is reverse transcription-polymerase chain reaction (RT-PCR). But it takes over eight hours to provide the result. Though rapid antigen tests give test results within 30 minutes, as per the Indian Council of Medical Research (ICMR), its false negativity rate is higher. With the ongoing unlocking of the national lockdown, the civic body needs more economical alternatives to rapidly and accurately diagnose cases of Covid-19. Thus, BMC has decided to start voice analysis to detect the virus within 30 seconds.

Dr Neelam Andrade, dean of Nair Dental Hospital and in-charge of NESCO, explained the process. When the symptoms of Covid-19 manifest, the person starts having breathing problems that affect the amount of air exhaled. Then the air interacts with inflamed muscles on its journey to production voices or speeches. These interactions impact the voice modulations — measurable qualities that form the basis of their biomarkers. Depending on its variations, an individual can be detected with Covid-19.

“There is one voice application that can be installed on mobiles or smart laptops. The suspected patient will be asked to count a few numbers in front of the device like a breath-analyser. The voice samples will automatically get synced with the main server of the provider. Then, through artificial intelligence, the result will be procured within 30 seconds,” she said.

The application will analyse voices of three types of people — suspected, confirmed and negative with Covid-19. Depending on their vocal biomarkers (VB), they will be diagnosed. For example, if a person’s VB is below the standard point — 0.5, he/she will be considered as negative. But if anyone records VB above it, he/she will be suspected with Covid-19 and an RT-PCR test will be conducted for confirmation.