From analysing how the virus has spread, to speeding up the drug discovery process, AI (artificial intelligence) has numerous potential benefits in the continued fight against Covid-19. Dr Tim Guilliams, co-founder and CEO of UK biotech firm Healx, discusses the role it can play during and after the crisis.

The Covid-19 pandemic has introduced unprecedented challenges for nations, healthcare systems and communities across the globe.

In response, scientists, researchers and medical staff have banded together to understand the virus and find new ways of developing treatments.

One of these new tools being used to help is AI.

Over the past few years, AI has been increasingly used to solve some of healthcare’s most pressing issues – from unlocking the power of medical records, to relieving doctors and nurses from repetitive tasks, to enhancing surgical precision, and more.

Just last month, we saw the UK’s first ever rollout of an AI-based cancer screening programme, indicating a new recognition, and adoption, of AI in healthcare.

Now that Covid-19 has introduced new hurdles for the industry to tackle, AI is again being leveraged to support the fight against the pandemic in a variety of ways.

Three ways AI can help in the battle with Covid-19

Forecasting the virus’ spread

Spotting and following the spread of the virus is an essential part of combating it.

By understanding how a virus spreads and how it is transmitted, experts can begin to predict where it may head next, or which communities may be at risk.

That’s why we’ve seen massive investment in tracking and tracing tools in the last few months – including from our own government in the UK.

But, as early as last December, some organisations were already sounding the alarm about the oncoming pandemic.

Indeed, in December 2019, a website called HealthMap, run by Boston Children’s Hospital, flagged a new type of pneumonia affecting people in Wuhan, China.

The website uses AI to scan data streams from social media, internet searches, news reports and more to identify signs of new outbreaks and was quickly alerted to the Covid-19 pandemic via a Chinese news outlet.

A similar approach can be seen with Canadian monitoring platform BlueDot, which used a data-mining algorithm to spot the outbreak and quickly shared information with its customers before the WHO (World Health Organisation) had even declared that it was a pandemic.

The company did the same thing back in 2016, and successfully pinpointed the location of an outbreak of Zika virus in South Florida.

Taking the strain off doctors and nurses

Several companies working within the healthcare sector have also deployed AI to alleviate the strain on hospital staff and help those who need medical attention be treated more quickly.

For example, Chinese firms have been using drones and AI-enabled robots to spray public areas with disinfectants, and dispense food and medicine to hospital patients, to help minimise exposure to doctors and nurses and ultimately lower the risk of infection.

Similarly, in an effort to address the huge demand placed on Covid-19 phone hotlines in Boston, Providence St Joseph Health system built an online screening tool to rapidly differentiate between those who may be suffering badly from the virus, and those who appeared to have milder symptoms or other ailments.

Within a week, Providence’s tool served more than 40,000 patients, using AI to deliver care and self-help to patients at an unprecedented scale, thereby helping to relieve some of the pressures placed on doctors and nurses.

This approach has been echoed in the United Kingdom, with tools like Doctorlink being used across the NHS to triage patients and provide online consultations.

The race to find a Covid-19 cure

It is no surprise that specialists are now also turning to AI to find a treatment for Covid-19.

With the drug discovery process taking years and costing millions, AI is being used to help accelerate timelines and reduce costs.

An innovative strategy for finding treatments for diseases quickly and safely is through repurposing existing approved drugs and identifying new uses for them.

This is our approach at Healx. Our AI platform, Healnet, is primarily used to find novel treatments for rare and often overlooked conditions – 95% of which still lack an approved treatment.

But we have also repurposed its power to assist in the hunt for Covid-19 treatments – exploring various drug combinations from approved therapies that could be used to combat the disease.

Discovering a potential treatment from combination therapies for Covid-19 is a process which requires detailed analysis of the eight million possible pairs, and 10.5 billion drug triples stemming from the 4,000 approved drugs already on the market.

This approach can be complex and lengthy, but our AI platform can overcome these challenges – synthesising vast amounts of data to quickly predict, and rank, combination therapies for our team of drug discovery experts to then interrogate and take to clinic.

What next for AI in healthcare?

Covid-19 has given us all new challenges to overcome, and new solutions to build. Whilst we’re still nowhere near to the end of this battle, AI has enabled us to innovate and fight the pandemic on multiple fronts.

What’s clear though, is that AI works best when it is paired with human insight, experience and empathy.

After the pandemic has abated, I hope to see more of this collaboration to find new ways to advance health care for the benefit of patients across the globe.