The origins of artificial intelligence can be traced back to the 1950s when Alan Turing questioned whether machines had the ability to think. It found its way into medical research in the 1970s where it was incorporated into a clinical setting. As scientists explore the use of AI, the number of misconceptions increase and we start to question whether its use is actually feasible or not. However, the current pandemic has shown that AI is a form of innovative technology that can be helpful in fighting Covid-19. It not only allows for correct screening and early detection of infection but it has major applications within the development of vaccines and medicines, all of which have positive implications on different sectors in society. But to what extent is its use feasible?
Throughout the pandemic, one of the problems faced by countries worldwide has been deciding how and when to reopen borders to foreign travel. This was a complicated matter since the capacity to test all foreign travellers was limited. Instead of testing a sample of visitors, Greece used a method that was rooted in artificial intelligence. With the input of Kimon Drakopoulos, a data science researcher who works at the University of Southern California, a machine learning algorithm called Eva was set in motion.
So, how does this method compare with regular testing? Although both methods are selective in the people that are being tested, the authors found the machine learning algorithm to be more effective when it came to identifying those that were asymptomatic as opposed to random testing or only testing those that came from a country of high risk. It detected two to four times more cases of infection than regular testing. This is because regular testing focuses on testing those that come under a certain risk category or it does so randomly. Comparatively, the artificial intelligence algorithm used the individual’s travel history and demographic data including age and sex. This allowed it to match the individual’s characteristics with previous data and thus more accurately estimate the risk of infection.
It detected two to four times more cases of infection than regular testing
Artificial intelligence has also paved the way for scientists to tackle the problem of analysing viral mutations. The Viterbi School of Engineering created an artificial intelligence framework that quickly analyses Covid vaccine candidates, finding the most effective medical therapies. When applied to the virus, the algorithm eliminated 95% of compounds that had the potential to treat the pathogen and located the better options. An artificial intelligence model such as this one can create vaccine design cycles that would typically take years in a matter of minutes. It predicted 26 possibilities, helping scientists to identify the best 11 to construct a multi-epitope vaccine: the greatest way to subdue the virus. At this time where virologists are concerned that the emerging mutations will evolve past the first vaccines, bringing existing vaccine efficiencies into question, this innovation could be invaluable in preventing the spread of the virus.
But are there any problems associated with the use of AI? When it comes to using AI to track the spread of viruses, large amounts of data are required – Big Data. This means that ensuring an individual’s privacy is protected becomes a more complicated matter. In order to track the spread of the virus, AI can be utilised to monitor individuals across multiple devices. Therefore, even though personal data is anonymised, as it becomes part of a large data set, artificial intelligence algorithms can de-anonymise these data through making inferences and deductions from other devices. For this reason, companies and the government require agreed systems to be put in place before relevant data can be shared with scientists.
Several outbreaks of viruses have marked the past two decades and these past two years have been no different. We can now see how AI can be harnessed in the prediction, tracking, and prevention of the virus. Leveraging Big Data and AI proves to be a powerful tool, playing a vital role in the fight against Covid-19. However, the methods of consensually obtaining one’s personal data and storing it securely are not yet clear, and so the use of artificial intelligence still proves to be an ethical issue.