March 18, 2020 – Before the world was even conscious of the threat posed by COVID-19, artificial intelligence had detected the beginnings of the outbreak.
On December 30, 2019, researchers from BlueDot, a company that uses AI to track and anticipate infectious diseases, spotted a report about a pneumonia of unknown aetiology in China.
Nearly a week later, on January 5, 2020, the World Health Organization (WHO) issued a request for more information from Chinese public health authorities. At the time, there had been 44 cases reported, with eleven patients in critical condition.
In the worldwide crisis that has since ensued, leaders from all sectors of the healthcare industry have turned to novel technologies to help monitor and control the spread of COVID-19 – the same technologies that were able to detect the initial outbreak.
Entities have leveraged artificial intelligence platforms, analytics algorithms, and data visualization tools to try to get ahead of the virus, or at least keep up with it. These technologies have the potential to anticipate where the disease will go next, as well as identify drugs that may be effective against COVID-19.
However, the issues surrounding AI in healthcare haven’t just disappeared in the wake of a global pandemic. Concerns about data quality, the absence of humans, and the overall accuracy of AI tools are perhaps even more valid now than they were before.
How are organizations currently using AI to combat COVID-19, and is the technology mature enough to handle a widespread health crisis?
AI’s current role in the global pandemic
As the world begins to take more and more precautions to reduce the spread of COVID-19, organizations have sought to get a handle on the virus using innovative tools.
On March 5, 2020, Google’s DeepMind published research discussing how they used deep learning to predict the structure of proteins associated with SARS-CoV-2, the virus that causes COVID-19.
“We emphasize that these structure predictions have not been experimentally verified, but hope they may contribute to the scientific community’s interrogation of how the virus functions, and serve as a hypothesis generation platform for future experimental work in developing therapeutics,” DeepMind researchers said.
At the federal level, policymakers are seeking expertise from the tech community to uncover new insights about COVID-19. The White House Office of Science and Technology Policy recently issued a call to action for AI developers to build tools that can be applied to a new COVID-19 dataset.
Community-based organizations are also leveraging AI to reduce the impact of the disease. At Medical Home Network (MHN), a Chicago-based nonprofit, leaders have implemented an AI platform to identify Medicaid patients most at risk from COVID-19.
“Community preparedness is critical. The dedicated MHN care teams who proactively reach out to patients via phone, email or text will help individuals understand what they can do to lower their chance of infection, recognize symptoms of infection and how best to access the advice and care they need,” said Dr. Art Jones, Chief Medical Officer at Medical Home Network.
To track the spread of the virus, several entities have employed data visualization platforms. On January 22, 2020, Johns Hopkins University first shared an interactive dashboard that tracks real-time data on confirmed coronavirus cases, recoveries, and deaths.
Additionally, officials in Latin America have rolled out a platform that facilitates the collection, investigation, and visualization of case and contact data during an outbreak.
Fewer humans could lead to more problems
Although AI has played a considerably big part in the COVID-19 pandemic so far, several concerns about the utility of the technology remain, particularly about its ability to function without human involvement.
With the US cracking down on social distancing, some major websites are turning to AI systems as they tell their employees to work from home – including Facebook, Twitter, and YouTube, BBC News reports. As a result, these tech giants are relying on automated algorithms to find and remove problematic material on their platforms, which the companies have admitted could lead to some mistakes.
While this development doesn’t directly apply to healthcare, it does raise some important questions about the ability of AI to stand on its own.
Patients, providers, and payers have previously expressed concerns about AI replacing doctors, but these anxieties have abated in recent years as most in the industry have come to view AI as a technology that can augment physician practices.
In the case of COVID-19, it seems that experts are taking a similar stance.
In an article published in The Lancet Digital Health, David Heymann, Executive Director of the Communicable Diseases Cluster, WHO, noted that people working at outbreak sites can collect vital data on COVID-19, which could then be used to train AI models. This data includes transmissibility, risk factors, incubation period, and mortality rate.
“We can’t replace the human brain at this point, nor the epidemiologist or virologist with anything that can analyze and rapidly do what is necessary at the onset of an outbreak. We still need to prime that AI with information from study of the evidence and link this to events in the outbreak,” Heymann said.
The ongoing concern over data quality
In addition to ensuring humans are a part of the disease tracking process, data quality issues are also top of mind for healthcare leaders seeking to use AI to combat COVID-19.
The industry’s persistent challenges with data volume, quality, and accuracy have been well-documented, especially as health systems move to adopt AI and analytics tools.
In the current climate, access to quality, accurate, and complete data will have a significant impact on whether AI tools will help or hinder COVID-19 prevention and treatment efforts. The number of confirmed cases increases daily around the world, as do the number of deaths. Keeping track of this information is as difficult as it is crucial, and leaders must make sure they’re training AI tools on the most up-to-date data available.
“Our data won’t necessarily come from China because it hasn’t been able to get hold of the data it needs because of the disorder and panic,” Heymann said in The Lancet article.
“This virus has spread to 24 other countries and these countries have set up extremely good systems of contact-tracing and patient isolation. This is where our information will come from.”
Moritz Kraemer, a spatial epidemiologist at the University of Oxford in the UK, told The Lancet that his team receives news reports and government reports twice per day that provide information on how many cases exist in any one location. Search engines and social media have also helped them track the disease.
“Before Chinese New Year we looked at how many people left Wuhan over a day, and this information comes from search engines. WeChat, a messaging, social media and mobile payment app, provides data on travel around Wuhan,” he said.
“Machine learning models use these data to predict the most likely location of where novel coronavirus might arrive next and this might inform where and how to run border checks.”
With all the uncertainty surrounding COVID-19 and the healthcare industry as a whole, it’s difficult to know how AI will ultimately fit into the pandemic. As the situation continues to evolve, so too will the technology. Whether it features large or small in the fight against this new threat remains to be seen.