Guest Column | August 1, 2018

Using Artificial Intelligence To Outsmart The Flu

By Jeff Heenan-Jalil, Wipro Limited


Flu vaccines are always less effective when compared to vaccines for other viruses; healthcare providers have long known this. However, the inoculations from this past season were among the least effective — on record at 17 percent effective against the most common flu strains. Treatments succumbed to a perfect storm of unpredictable mutations during its development and a particularly active hemagglutinin (HA) by the H3N2 virus.

But this is just a repeat — albeit an extreme one — of the same problem we face every year: an inability to accurately predict the virus’ next mutation. Think of viruses the same way you'd think of a persistent computer hacker. Every time the system is attacked, we can create anti-virus software to destroy it, but that next attack will always include some new code that penetrates the software's security measures. It's a seemingly endless game of both parties evolving and trying to one-up the other, and in this case, the flu's sporadic egg mutations puts the researchers and scientists developing inoculations at a disadvantage.

The efficacy of next year’s vaccines remains to be seen, but one thing we do know is that the flu's HA will continue to evolve. There must be a better solution to outsmart it, and a growing body of experts believes it's one that doesn’t rely on guesswork. The solution? Artificial intelligence.

Fighting Back

For decades, the World Health Organization chose the reference strains that built flu vaccinations. The cocktail of strains must be incubated in eggs, and mutations are necessary for them to grow. Many of these mutations are insignificant, but recent events have shown that many of them can influence the outcome of the entire flu season.

To eliminate the risk of mutations, researchers have developed several alternatives. For example, Flucelvax Quadrivalent and FluBlok are grown in canine kidney cells and caterpillar cells, respectively. FluBlok is specially made for people with food allergies, and because it is cell-based rather than egg-based, experts believe it may have been more effective than some previous vaccines.

The hope of a universal, multiyear flu vaccine, however, is growing and may come to fruition thanks to combined efforts by biotechnologists and academics. For instance, a program headed by the National Institute of Allergy and Infectious Diseases has changed how flu vaccinations defend against a virus. Rather than focusing on the head of the HA protein, the vaccine stimulates antibodies that bind to its stalk, which are common among most strains.

Other programs have utilized machine learning to more accurately track outbreak conditions at specific times and in precise locations. With the help of AI, researchers hope to one day pinpoint the exact effect certain mutations can have on how resistant the flu virus is to antibodies.

AI exists to aid humans in gathering, sifting through, and analyzing data to better understand patterns and predict future outcomes, and that's precisely what the health sector needs more of in combatting the flu. One virologist, Richard Webby, conducted a machine learning study on flu vaccines in 2009 that health scientists still rely on to better understand the virus's molecular structure.

In many ways, we’re learning to fight the influenza virus in much the same way we’ve learned to fight off hackers’ computer viruses: by identifying patterns and methods of attack and creating solutions against them. Using technology, we can understand and defend against the variations that make it so effective as well as predict which of those variations deserve the most focus.

Last flu season was an especially tough one — but with today's technology on our side, it may just be the turning point in our fight against the flu.

About The Author

Jeff Heenan-Jalil is the senior vice president and global head of Health Business for Wipro Limited, where he is responsible for the business unit’s profit and loss, strategy, and operations. Jeff has more than 22 years of global diversified experience acting as a change-agent for IT services businesses. Outside of Wipro, he is a board member of Digital Square, an organization that aims to empower countries to develop sustainable national digital infrastructures. In 2013, Jeff was placed among the Top 50 in the Global Telecoms Business — Power100 Rankings. The views expressed in this article are Jeff's, and his employer does not subscribe to the substance or veracity of Jeff's views.