The coronavirus is confusing for the common man, since its similar to any common flu. And this creates confusion about who can, or should, get tested. To tackle this problem, Sydney-based Medius Health has created Quro, a risk assessment tool to determine whether you should be tested for Covid-19 or not. The AI digital health startup run by Indian Abhi Bhatia has so far done more than 75,000 consultations since rolling out the tool on March 4 in India and Australia.
“When Covid-19 hit, we had enough understanding of how we can utilise our medical knowledge graph at the backend and developed an AI-powered risk assessment tool which presents the users with questions related to symptoms and risk factors in accordance with the World Health Organisation protocols and guidelines,” the founder of Medius Health told indianexpress.com over phone from Sydney.
Quro, an AI-based risk assessment tool, is like an anonymous consultation where people do not need to sign up or give personal details. The tool, which acts as a virtual health guide, asks a number of questions such as your travel history, past medical history, your symptoms, what sort of problems you are having right now from a symptomatic point of view and so on. The tool is free to use and can be accessed on the web.
But collecting the data is one part of the story, says Bhatia. His team, which consists of data scientists and clinical scientists, has created an artificial intelligence (AI) system called disease frequency and evoking strength that extracts the data and using different combinations, it understands the relation between symptoms and disease. And based on that, its AI system predicts whether a person is at higher risk for getting sick from Covid-19 or not.
The system gets updated frequently, based on the data and the trend provided by the World Health Organisation (WHO) and the Centers for Disease Control and Prevention (CDC). “We keep on updating the prevalence rate value of those relationships between coronavirus and a particular symptom. So for instance, when coronavirus started picking up, the doctors realised that nasal congestion is the one thing that actually differentiates between a flu as well as coronavirus, but then a lot of things have changed,” he said.
Bhatia said the aim was to “quell the panic” and at the same time offer “the right set of information” so that people should know that this could be a mere cold and flu and not Covid-19. He further added that the tool will help identify those people who have a high risk of Coronavirus and need to get the testing done.
The tool [Quro] will help people understand how to differentiate between symptoms that are more relevant to coronavirus versus influenza. That is the biggest problem right now because everyone who’s coughing and has some sort of nasal congestion, they feel that it’s a coronavirus and it’s creating a lot of panic among people,” he explains.
Citing the example of India, Bhatia said the risk assessment tool could be useful for a country with a population of 1 billion, where unnecessary fear among people about Covid-19 can put a lot of burden on hospitals and healthcare institutions.
According to Bhatia, 69.2 per cent of people from India who have accessed the risk indicator have lower chances of contracting Covid-19, 22.7 per cent of profiles fall in the moderate risk category and just 9.1 per cent of people are at a higher risk of complications from the deadly virus.
Bhatia feels that the tool can help government agencies and health organisations in understanding the magnitude of Covid 19 and take necessary steps before it gets too late to stop it from spreading.
The company has reached out to the Ministry of Health to push out the risk assessment tool to a large population base in India. Meanwhile, Medius Health is working with a couple of partners in India to deploy the tool, including ICICI Prudential and Bajaj Financial Services. The company is also developing a special platform for Max Healthcare.
It took Bhatia and his team 8-9 days to develop the tool. “Before we roll out any AI tool, we train the model and then we have to do internal testing. In this case, we created different kinds of case simulations that we picked up from Wuhan. For instance, who was traveling, where these people were traveling, what was the age group, what was the gender, what geographical patterns that they have, what was the onset of symptoms? So we picked up that data from WHO and created AI-driven case simulations that we then injected into our platform,” he said.
At a time when the developed countries like the US and Italy are failing to stop the coronavirus outbreak, Bhatia acknowledged that the use of Artificial intelligence (AI) can be a game-changer in healthcare.
“AI comes into play as early detection, collecting the data on a mass scale so that we can create epidemiological population-level information through which the government can allocate the resources at the right place, and put people in the right bucket. So if there’s a high-risk person, he should be the one who should be isolated first and get proper testing done,” Bhatia said.
He further that those data points will further help train models and the AI engine would even be able to predict within those high-risk people who are least touched to coronavirus and who are more prone to infectious disease.
Medius Health, which was founded in 2017, has a team of 32 people including data scientists, doctors and software engineers. The startup has so far raised AUD 4.2 million in two rounds and is looking to close its Series A from a consortium of investors from India and Singapore.
📣 The Indian Express is now on Telegram. Click here to join our channel (@indianexpress) and stay updated with the latest headlines