×

How Artificial Intelligence-backed technology is helping in the battle against COVID-19

Interview, Prashant Warier, CEO and founder, Qure.ai

Prashant Warier, CEO and founder, Qure.ai

COVID-19 has caused some irrevocable damages to the public healthcare systems in countries across the world. As scientists race against time to find the right vaccine and cure for the pandemic, technologists too have chipped in with solutions in the field of Artificial Intelligence (AI), offering assistance in testing, tracing and treatment.

A number of institutions and companies in India are coming up with indigenously developed, AI-backed technologies, which are likely to take the country and the world a step closer to tackling the virus. In an interview to THE WEEK, Prashant Warier, CEO and founder, Qure.ai, which is now deploying AI-powered solutions for COVID-19 management by using chest X-rays which are about 10 to 15 times cheaper than the RT-PCR tests, speaks about how technology is helping in the battle against the pandemic.

You use Artificial Intelligence to detect and triage COVID-19. Please elaborate on how it is done

COVID-19 ranges in severity—from asymptomatic, mild to severe. The virus is known to mutate and change, making it extremely hard to treat and contain. Having all information at hand and having actionable insights is something that will aid healthcare providers in offering better care. Today, the usage of chest X-rays is being recommended by the American College of Radiology, the Royal College of Radiology and the Canadian Association of Radiologists to screen for COVID-19. Many hospitals across the globe have already implemented chest X-rays in their COVID-19 diagnosis and progression protocols.

The lungs of a COVID-19 patient is very different from that of a healthy person. That is where Qure.ai’s technology comes into the picture. In 2017, we developed an AI technology that could identify abnormal findings on a chest X-ray. Research shows that in COVID-19 patients, lungs will have consolidations and ground-glass opacities, all of which are detected and marked out using our algorithm. The solution can also quantify the volume of infection, automating a time-consuming step for the already resource-strapped physicians. This is ideal to monitor ICU patients to understand if their lung condition is improving or worsening on a daily basis.

Our solution has also found use in locations where there are not enough swab test kits. The chest X-ray is about 10 to 15 times cheaper than RT-PCR tests, widely available and also can be done in a few minutes, compared to several hours taken for the RT-PCR tests. As such, they have been adopted by various hospital groups in triaging who should be further tested for COVID-19. Unfortunately, radiologist interpretations of the chest X-rays might take hours, sometimes days. That is where Qure comes in, by automatically processing a chest X-ray in under a minute and identifying the COVID risk level of the patient.

What has been the role of Qure.ai in the field of tuberculosis over the past years?

Tuberculosis (TB) is one of the oldest and most infectious diseases of all times. While it is totally curable, it is still one of the top 10 causes of deaths worldwide. In India and some of the emerging economies, there is a huge dearth of radiologists to read chest X-rays that are used as a screening for TB. In India alone, 80 million chest X-rays are being acquired every year. There aren’t enough radiologists to read them within an acceptable timeline. Depending on the availability of radiology expertise, it can take anywhere between one to 14 days to get diagnostic reports. This leads to delayed diagnosis of the TB patient and further spread of the disease. This gap can be minimised by automating and classifying chest X-ray readings as normal or abnormal with a solution that is scalable and requires little manual intervention. This is precisely how Qure’s solution, trained on more than 2.5 million chest X-rays, supports the grassroot level health infrastructure. The tool automates reading chest X-rays and generates reports within seconds, thereby reducing the waiting time for TB confirmatory tests, from weeks to a couple of hours and allowing for treatments to be started on the same day. Today, we are present in more than 40 sites across some of the highest TB-burdened countries and have touched more than 200,000 lives.

Can Artificial Intelligence backed technology by Qure be applied in high population density areas of Mumbai such as slums?

Yes, Qure is already working in Maharashtra with Municipal Corporation of Greater Mumbai (MCGM). Qure has partnered with MCGM and other private entities and launched the first mobile COVID-19 testing unit earlier this month.

The COVID-19 bus, equipped with Qure’s AI and a host of other medical testing facilities like oxygen saturation and body temperature checks, mass screens individuals in the high-risk areas of Worli. Densely populated Worli-Koliwada, one of the first containment zones in Mumbai, is a region of specific interest since it is a high-risk area, comprising of slums and fisherman colonies. The COVID-19 screening bus enables close contacts of positive patients and other asymptomatic residents to be screened via X-rays on the spot, along with checks for fever and oxygen saturation levels. Following the qXR analysis, medium and high- risk candidates are immediately directed for the next steps in the pandemic management process.

In the last 30 days, we have screened more than 2,000 cases and around eight per cent have been instructed for further medical attention. More such screening buses are in the pipeline for other high-risk areas in the city.

What have been your observations so far, with regard to disease progression, given that your technology has been used by health systems in India and abroad?

The progression monitoring capability of qXR is currently being used in San Raffaele Hospital in Milan, Italy, and at the NHS with Royal Bolton Hospital in the UK. They have had very positive feedback for us since we are now supplementing their reports with the ability to give a quantified report, something which was previously rare for chest X-rays. This adds a lot of value in their workflow.

According to the Bolton Critical care team leader: “From my experience, the implementation of this innovative solution has been very positive, I am still getting our junior team used to looking at the additional information the software provides. The software has identified successfully a range of pathology from different patterns of consolidation to pneumothorax. I think a significant

benefit will be seen in the post COVID era when we return to our more usual range of clinical presentations. I’m hoping this will improve the diagnostic accuracy when our junior doctors first assess patients, and should also improve the correct coding and identification of community-acquired pneumonia.”