Hi All,

We need to help Indian Researchers getting more voice samples for their research project COSWARA. Read the entire article.

Due to the SARSCov2 Pandemic the entire world is trying their best to understand, diagnose, cure the disease. The scientific world is putting lot of efforts and many are working literally day and night to find, develop technologies that can help the humanity get through the pandemic crisis.

I read a lot about passive voice sampling and there was one particular Israel company MyEleanor, a voice bot and virtual care manager from MyndYou, can call individuals or act as a hotline to assess risk, manage symptoms and provide guidance. On each call, AI-driven voice analytics can detect subtle changes in health and trigger proactive interventions.

One such effort is also ongoing in the Bengaluru-based Indian Institute of Science (IISc). A group of researchers are trying to evaluate the possibility of COVID-19 diagnosis from just the respiratory sounds of the person. A paper documenting the early efforts by the group is posted as PDF

There was also an article in Hindustan times, explaining their efforts. But here we are going to focus on how we as Indian citizens can help these researchers in their goals.

Coswara is an attempt to provide a simple and cost-effective tool for diagnosis of COVID-19 using breath, cough, and speech sounds. As most of the major symptoms of the disease include respiratory problems, the proposed project aims to detect and quantify the biomarkers of the disease by mathematical analysis of these sound samples of an individual.

More the samples more the accuracy! visit https://coswara.iisc.ac.in/

Researchers are actively looking for voice samples from healthy individuals, individuals suffering from respiratory aliments like Asthma, Chronic Obstructive Pulmonary Disease (COPD), Chronic Bronchitis, Pneumonia, Pleural Effusion, Bronchitis, Emphysema, Cystic Fibrosis, Bronchiectasis,  etc and Covid-19 positive patients.

What happens to the donated respiratory sound samples?

The curated dataset is released as open access via Github. As the pandemic is evolving, the data collection and analysis is a work in progress. The collected data will be analyzed using signal processing and machine learning techniques. The goal is to build mathematical models aiding the identification of COVID-19 biomarkers from sound samples.

Diagnostic tool

The final goal of the project is to release the diagnosis tool as a web/mobile application. The application will prompt the user to record respiratory sound samples, similar to the dataset collection stage, and will provide a score indicating the probability of COVID-19 infection. The final deployment of the tool is subject to validation with clinical findings, and authorization/approval from competent healthcare authorities

In addition to IISc, at least three research groups at Carnegie Mellon University, in the United States of America, Cambridge University in the United Kingdom, and École Polytechnique Federale de Lausanne in Switzerland are also working on similar COVID-19 pre-screening diagnostic tools.

“We have received a good number of respiratory sound samples donated by healthy participants. It has been great to see people supporting the idea. However, we also need to make the project really popular among individuals with respiratory ailments and COVID-19 patients. This has been difficult so far may be because we have not been able to reach them.” Neeraj Sharma, one of the contributor to this tool, said.

” We have made our project website easy to use, the whole process of donating your respiratory sound samples just takes 5-7 mins, and can be done from any quite place (like your room) and with your personal mobile phone.”

He further adds,” We are putting efforts to contact hospital authorities who can spread the word about the project among the patients. Once we get the data collection gain good momentum, it will start getting clear how effective this idea can be in reality. Recently, we presented the idea to ICMR, and they gave us a nod to ramp up the data collection, sooner the better.”

We need to help these researchers in doing better by contributing our voice samples (takes 5-7 mins). No personal identification information is collected. Share this with your medical care providers, friends, relatives and contacts which will help them reach the larger population.

I am sending the link of this article to some physicians that I know could take this forward and reach the positive cases for voice samples.

Together we will be able to sustain through this phase.

Stay Home Stay Safe

Amaraja