The accuracy required to get FDA clearance for multiple findings is much higher than the burden of proof for one finding at a time, says Prashant Warier, Qure’s co-founder and CEO
When a patient comes to the hospital emergency room with a stroke or head injury, every minute counts. CT scans are typically used to identify the problem and decide on appropriate treatment. Radiologists interpret scans but they’re in short supply, often leading to life-threatening delays.
Mumbai-based startup Qure uses an AI imaging tool qER to save precious minutes for emergency room staff to take action based on head CT scans. After deployment in India and several other countries, qER is now entering the US where 75 million CT scans are performed every year. A couple of weeks ago, Qure received US FDA 510 (k) clearance for this product. What makes it special is a four-in-one clearance. The tool has been cleared for triaging four critical conditions—intracranial bleeds, mass effect (due to spaces in the brain filling up), midline shift (in the brain’s alignment), and cranial fractures. “The accuracy required to get FDA clearance for multiple findings is much higher than the burden of proof for one finding at a time,” says Prashant Warier, Qure’s co-founder and CEO. AI tools such as qER are coming into play due to the effects of covid-19. Scans go to a teleradiology centre for reading by radiologists working from home. AI can not only speed up reading a scan, but also prioritise scans that indicate something critical like bleeding in the brain. Those go to the top of the radiologist’s stack of scans.
It took Qure a year-and-a-half to get the FDA clearance. Even now, it’s only for a triage solution—that is, qER is allowed to suggest what a radiologist should read first. For an AI-based diagnosis on par with that of a radiologist, it would require further studies. One of the challenges for healthtech startups outside the US is that FDA insists on validation based on US datasets. So Qure had to source data from the US before conducting tests validating qER as a triage tool. “Nobody in the organisation had gone through FDA approval before, so we relied on consultants. Now that we have gone through it, the next approvals will be easier because we have contacts with US hospitals and radiologists for the studies,” says Warier. Qure’s AI tools have been commercially deployed in Asia as well as Europe, where it has CE certification. A new use case in the post-pandemic scenario is reading X-rays to detect covid-19. Qure’s other product, qXR, can detect 24 types of abnormality from chest X-rays. It re-purposed this product in March as a supplement to covid-19 tests that require swabs.
The RT-PCR (reverse transcription polymerase chain reaction) tests are time-consuming and costly. The humble X-ray has emerged as a quick triaging tool, so that those showing signs of covid-19 in the lungs can be prioritized for RT-PCR confirmatory tests and treatment. The AI-based qXR speeds up and expands access to X-ray triaging. There was insufficient data to build a covid model at the outset. So the Qure team looked up publications from China which detailed how covid-19 shows up in lungs. “We took our existing models which can detect opacities and other abnormalities, then we put a layer on top of that to distinguish covid-19 from other lung conditions,” explains Warier. Hospitals in the UK, Italy and Mexico have deployed qXR for triaging covid-19, apart from multiple sites in India, Pakistan and the Philippines where mobile vans with X-ray machines are going into densely populated areas. When Qure started out in 2106, it built qXR as an AI tool to detect TB from chest X-rays. This too involved outreach with vans. And over time, after ingesting data from millions of X-rays, it evolved capabilities to detect various other lung diseases such as pneumonia. The qER product followed and more are on the anvil.
Sequoia India led a $16 million funding round for Qure in February this year, making it one of the top three funded Indian startups in healthtech AI. “The quality of healthtech products being built out of India, and quality of teams that are building these products is at par with their global peers,” says Anjana Sasidharan, principal, Sequoia Capital India Advisors. Armed with a PhD from Georgia Tech, Warier cut his teeth on data science for logistics and ecommerce before plunging into healthcare AI. His co-founder Pooja Rao has a PhD in neuroscience and combines that with data science. They identified healthcare as an area that was lagging in adoption of AI. They’re out to change that with deployment of their AI tools globally.