Knowledge is Power
The outbreak of COVID-19 has changed the very fabric of our culture. Billions of people worldwide are under lockdown, while many businesses and employees face an uncertain economic future. This time, the world was unprepared for Coronavirus and is suffering the consequences.
However, one good thing has emerged from this pandemic: COVID-19 has accelerated the innovation and development of the AI systems that may prevent the next pandemic from taking place. Next time there is an outbreak, we will be better prepared.
Just before midnight on December 30, 2019, an artificial intelligence company called BlueDot flagged a news report from China concerning a cluster of “unusual pneumonia” cases occurring around a market in Wuhan, China.
BlueDot, which constantly sifts through hundreds of thousands of online sources in 65 languages, alerted its human colleagues who immediately recognized similarities to the SARS outbreak in 2003. Nearly a week before the World Health Organization (WHO) or Chinese authorities released statements on the emergence of the Coronavirus, BlueDot had already spotted it.
And if action had been taken sooner, the world might be looking very different today. “If health officials could have taken action earlier and contained the outbreak in Wuhan, where the first cases were reported, the global clampdown could have been at a much more local level,” said Professor Richard Kuhn to Time Magazine.
BlueDot was created by epidemiologist, physician and professor Dr Kamran Khan, who witnessed first-hand the devastation of the SARS outbreak. He wanted to create a company that would “spread knowledge faster than diseases spread themselves”.
Khan knew the importance of timely information. In an interview with CNBC, he revealed that doctors don’t get information about infectious diseases from government agencies fast enough. Instead, epidemiologists like him often use the internet to gather real-time information. Khan realized that an AI-enabled algorithm would do a far better job at sorting through data than humans. AI can operate nonstop and can process incredible amounts of data in a much faster time than humans.
In fact, the speed and scale of AI is unrivalled. “When it comes to detecting outbreaks, machines can be trained to process vast amounts of data in the same way a human expert would,” explained BlueDot creator, Dr Khan.
AI algorithms like BlueDot can mine billions of news reports and posts from around the world, helping experts to detect the early stages of a potential outbreak, and stop an outbreak from becoming a pandemic.
Communicating valuable information to the public is crucial to containing the spread of a virus. The World Health Organization, for example, has partnered with WhatsApp and Facebook to create an AI-enabled chatbot that will provide vital information to users through their smartphones. The chatbot will answer common questions related to the pandemic, provide the latest news, and teach users how to protect themselves and others.
Going forward, governments can use AI to track public sentiment, which can provide officials with insight into what people are thinking or saying about the disease. Communication can then be tailored to this sentiment, effectively educating the public when they are wrong, or reinforcing information that can help stop the spread of the disease.
Korean scientists have used a deep learning-based model to help find an effective vaccine. Coronaviruses invade human cells through a “spike” protein, which binds to a human cell, allowing it to infect said cell.
By binding the spike protein of a coronavirus with another molecule (antibodies), scientists can prevent the virus from infecting human cells. The deep-learning model created by the scientists was able to predict “how strongly a molecule of interest will bind to a target protein.” From these predictions, the model detected four antivirals that could potentially bind COVID-19.
There are several companies engaged in the race to find an effective vaccine against COVID-19. Google’s DeepMind, for example, is using deep learning to gather new information about the structure of proteins that will form the basis of a vaccine.
On their website, DeepMind researchers had the following to say:
Other scientists are using deep learning algorithms to accurately detect COVID-19 from CT scans of the chest, while yet more have developed a mobile digital X-ray system that uses AI cloud-based software to help with early diagnosis. The company that developed this system plans on designing the machine so that it can be installed in public places like train stations and airports. Meanwhile, Google’s Verily is in the process of developing a temperature patch that transmits data to a phone app, alerting its users to higher body temperatures.
Facebook is sharing anonymized data with researchers from American and Taiwanese universities to track movements to forecast the spread of the virus. Although controversial, geotracking can help to effectively trace the location of patients, and the people they have come into contact with. According to Time Magazine, this strategy worked well in South Korea and allowed officials to more easily identify the contacts of members of a church that formed the first major cluster of infections.
AI can be used to power robotic cleaners that can clean hospitals, reducing the risk of exposing humans to the virus. In China, navigational AI algorithms are helping ambulances reach patients faster, while robots are being used in public spaces to warn people who aren’t wearing face-masks.
AI is our most effective weapon
Combining human intelligence with the power of machine learning will no doubt prove invaluable in the fight against pandemics like the one we are experiencing. Besides finding a vaccine in a much faster time than humans could, AI could become hugely important in detecting outbreaks before they have a chance to spread. It’s long been recognized that AI has a crucial role to play in healthcare, and our current crisis only serves to strengthen this belief.