HEALTH | 04.02.2020
Artificial intelligence can help us combat COVID-19
Major pandemics impacted humankind in centuries gone by and the scourge continues in the 21st century with the outbreak of COVID-19.
Experts and researchers from around the world are joining forces and resources to fight this coronavirus and new threats from viruses using one of the most powerful advances in technology—artificial intelligence (AI). There are currently many projects based on this form of technology, all focused on one thing—containing COVID-19 and discovering its impact.
Thanks to AI, a Canadian company, which specializes in tracking and anticipating the spread of infectious diseases, detected COVID-19 well before it spread beyond China. With machine learning algorithms and big data, its BlueDot software warned on December 31, 2019 of certain cases of “unusual pneumonia” around a market in Wuhan, China. It was not the first time it had anticipated a catastrophe. In 2016, this Toronto-based startup predicted the Zika outbreak in Florida six months before it occurred.
To make its predictions, BlueDot uses natural language processing and machine learning. By means of these processes, it selects official data from the WHO and various other sources. These include reports and online articles by healthcare specialists and journalists, as well as data on climate change and airline tickets issued around the world. According to the website, “our platform scans over 100,000 official and mass media sources in 65 languages per day.” https://bluedot.global/
In terms of the predictions for China, two sources of data were key: Chinese articles which reported 27 cases of pneumonia in a market selling seafood and live animals in Wuhan, and data on the movement of those potentially infected. This led it to anticipate and publish the first scientific article on COVID-19, in which it warned of global spread.
AI—60 times faster than humans
Artificial intelligence or deep learning has been shown to be as effective as doctors for diagnosing diseases, based on the analysis of images. In the case of the coronavirus, the research institute Damo Academy, https://damo.alibaba.com/, which belongs to the Chinese giant Alibaba, has developed an algorithm capable of identifying new cases in 20 seconds with an accuracy of 96 percent, based on CT images. In order to train this diagnostic tool, researchers used sample data from more than 5,000 confirmed cases. According to researchers, this tool is approximately 60 times faster than human detection.
A similar tool for the same purpose, developed by the Huazhong University of Science and Technology, alongside the companies Huawei and Lanwon, not only helps doctors with rapid diagnosis, but also helps to reveal how COVID-19 behaves at different stages of the illness, as well as the effectiveness of drug therapy.
Working together to find a solution
While waiting for a vaccine, researchers and doctors around the world are working to find a treatment that cures the coronavirus or slows its spread. To do this, companies such as Google, with its AlphaFold system, are providing the scientific community with all the information they have from studying the proteins that make up the virus.
The search giant is not alone in this task. The information obtained from RNA structure prediction algorithm (LinearFold) from the Chinese company Baidu http://research.baidu.com/Index is also helping efforts to try to find out how the virus works and develop a vaccine.
While collaboration between the technology sector, healthcare sector, research centers, organizations and governments is more transparent than ever at a global level and is growing by the day, there is still a long way to go and much work to be done. The solution to the current crisis will help us deal better with future pandemics.