No doubt, most of us have already heard of the Coronavirus outbreak from Wuhan, China, since it has already spread to over 10 countries. Earlier this month, officials from groups like The World Health Organization ( WHO ) and the ( USCDC ) broke news about the severity of this outbreak. But did you know that an AI system actually predicted this days before any organization said what they think publicly? You heard that right. AI is being used to predict disease outbreaks throughout the world. So let’s take a real quick look at this technology and the company driving it.
By December 31st, 2019, this AI tech developed by BlueDot was able to predict the outbreak nearly a weekly before the American CDC and the WHO announced it. Their AI-driven algorithm was able to do this by scanning the internet for news reports in multiple languages, and global airline ticketing data to turn out accurate predictions for the location of the outbreak. BlueDot even correctly predicted that the virus would head from Wuhan to Bangkok, to Seoul, Taipei and Tokyo. And just to clarify, the AI algorithm isn’t fully responsible for these discoveries.
When it comes to any AI-based disease outbreak discovery, human intervention plays a key role, since all data must be interpreted by scientists to make sure it makes sense. And once the findings are double-checked, BlueDot sends a notification to it’s clients, who are usually hospitals and governments. But just to be clear, the average person can’t sign up for their services just yet. It’s intended for established organizations and governments. For example, the CDC has partnered with BlueDot in the past, when analyzing and releasing information about diseases, but let’s look at the bigger picture.
This kind of technology in the healthcare field could be life-changing by facilitating faster outbreak predictions. Just look at BlueDot’s track record. In 2016, it predicted the Zika virus outbreak in Florida six months before it happened. And now the company has raised millions of dollars to see what it can do at a higher capacity.
Still AI prediction technology for diseases has been around for years. So why does it seem like it’s just now coming to surface? Well, let’s go back to 2013, when Google Flu Trends was shut down. Before this Google project was terminated, it had been running since 2008 and said that it could predict disease outbreaks based on people’s Google searches. The idea behind it was that if people had flu symptoms, then chances are, their online searches, like looking for the nearest doctor, could predict an outbreak. But the program was shuttered in 2013, because it underestimated the severity of that year’s flu season by 140%.
But why does BlueDot seem to succeed where Google failed? It seems to be because BlueDot is relying on more specific information than just searching queries. According to their founder and CEO, we can pick up news of possible outbreaks, little murmurs or forums or blogs of indications of some kind of unusual events going on. By searching for specific words and phrases in over 65 languages, and across news outlets and other online sources, their algorithm was able to get an idea of when an outbreak is just emerging. Although he made it clear they don’t just rely on everything posted on the web, saying that their algorithm ignores social media posts, because they can be unreliable.
On top of just outbreak predictions, AI could be used in a broader scope in the healthcare field. Last year, IBM came up with an AI model to predict the development of breast cancer with the near same level of accuracy as actual radiologists. By using imaging data from mammograms and patient health history, the system was able to correctly predict the development of breast cancer in over 80% of cases. So now there’s reason to think that AI will play a more crucial role in healthcare in the future. And maybe to the point that it will save your life.
So that being said, we are going to leave some questions for you. Do you trust AI systems to accurately predict outbreaks? BlueDot is one example, but as we saw, there have been past failures. On the other side of that argument, do you think this technology needs to have more money invested in it, since it has delivered some promising results? Leave a comment down below and let us know what you think. If you like this article and you want more, be sure to give it a thumbs up and fill that subscribe form.
