How a Mumbai start-up is helping Italian doctors test Covid-19 patients

Qure.ai's AI-backed solution interprets X-rays to help doctors classify patients as high-, medium- or low-risk and prioritise testing accordingly

With the number of Covid-19 patients across the globe having crossed the 12-million mark, nations are grappling with a chronic shortage of test kits.

Apart from the production of more such kits, another solution to this problem is triaging. In medical terms, it is the assignment of degrees of urgency to wounds or illnesses to decide the order of treatment of a large number of patients. This would help find out who needs to be tested on priority.

A Mumbai-based start-up realised in February that Covid-19 is going to hit the nation hard and gathered all resources to build solutions to fight the pandemic.

Qure.ai already has an AI-backed solution called qXR, which can automatically generate chest X-ray interpretation reports, detect tuberculosis, chronic obstructive pulmonary diseases (COPD), lung malignancies and medical emergencies such as lung collapses and cardiac disorders.

“We have re-purposed qXR to additionally interpret an X-ray to detect findings indicative of Covid-19 via a Covid-19 score,” says Prashant Warier. This solution is now being used to monitor disease progression in Covid-19 patients in hospitals across six countries, including Italy's San Raffaele University Hospital in Milan.

An R&D team of 10 people studied data available from China which showed that Covid-19 infection is present in the lungs as bilateral opacity. These are basically white spots on both sides of the lungs and are present as a peripheral distribution. Based on the bilateral and peripheral nature of the spots, the size of the infection and the opacity, the team created an AI-based algorithm called the Covid-19 score. 

Looking at a Covid-19 score, a health professional can classify patients as high-, medium- or low-risk. “This can help identify triage patients who need to be tested further for the virus using an RT-PCR test. As many nations are seeing a shortage of test kits, the score can prioritise those who need to be tested and those who need to be asked to self-isolate,” explains Warier.

This process is also helping Italian health workers quantify exactly how much of the lung is affected and how the patient is progressing with therapy. The solution, developed two weeks ago, has been used on 10,000 patients so far.

Sequoia-backed Qure.ai has developed another solution called qScout for frontline health care providers for contact tracing and remote triaging of patients. 

How it works: A use case

After the lockdown, if someone is coming from a foreign nation, his information and that of his primary contacts is registered on the portal by airport authorities. The person's condition can be remotely monitored using a chatbot that works on WhatsApp.

If by chance the person tests positive later, all his primary contacts would be notified within 2-3 minutes on WhatsApp and alerted for symptoms such as cold, cough and breathlessness.