COVID-19. Data and Artificial Intelligence in the fight against the pandemic

The 2019 coronavirus disease (COVID-19) pandemic has unleashed one of the most severe global crises of recent times. This is due to the dizzying rate with which it spreads and the high danger it poses to people with some condition vulnerable to the acute respiratory syndrome it causes.

Within 71 days after the first case was confirmed in China (I write this on March 25), the disease has spread to 170 countries and territories with more than 467,000 confirmed cases and 21,000 deaths. The number of cases took 53 days to reach 100,000, 12 to get to 200,000 and 6 to exceed 400,000.


Prepared with data from the World Health Organization

In some countries, the case fatality rate (percentage of deaths in total confirmed cases) is much higher than the maximum rate observed in China since the first case was confirmed, which magnifies the severity of the effects of the disease in those countries.

CFR_20-mar-25pngPrepared with data from the World Health Organization

In order to mitigate the spreading, some countries have implemented drastic measures of social distancing and isolation, which practically cause the suspension of daily activities. Unfortunately, such measures will bring very serious impacts on the economy and for the lives of millions of people, from which it will take months or even years to recover.

The spread and danger of COVID-19 are favored by an interconnected world and, paradoxically, by higher life expectancy. According to the International Air Transport Association (IATA), 22 thousand pairs of cities around the world are connected by scheduled flights that transport 4.3 billion people a year (11. 8 million per day). On the other hand, although not at the same rate in all countries, during the last 50 years the life expectancy at birth increased from 57 to 72 years, although sometimes advanced age is accompanied by chronic conditions.

This pandemic presents itself then in unique historical circumstances, which give it a greater scope and impact than any other similar crisis. However, it also comes at a time when high-capacity, high-speed communication networks, as well as computers and mobile devices with capacity unimaginable a few years ago, resulted in the era of hyper connectivity and the rise of data science and artificial intelligence (AI). These circumstances have the potential to make a difference in our favor in the fight against this threat.

The most notorious effect of these favorable circumstances is that we have information on the global evolution of the pandemic in near real time. In addition to the human note of the media and social networks, through which we witness how the crisis is lived in other parts of the world, it is easy for anyone to get very complete data from the same sources that feed the media and governments. Three of the main sources available are:

  • The website of the World Health Organization, which contains medical and guidance information, as well as statistics from around the world that are updated every 24 hours.
  • The John Hopkins University dashboard, which updates global information every hour, and from which databases can also be downloaded.
  • The website of the Mexican Ministry of Health, or the corresponding health authority of your country, that provides information of national interest and educational material.

This high availability of information has encouraged researchers and practitioners of data science and AI from across the world to conduct and publish research on various topics related to the pandemic, to educate public opinion and contribute to the fight against the threat. Obviously, not all articles have the technical rigor and quality to make them worth reading, but we can select them based on the profile of the author and the ratings it has on the site that publishes them.

An example is an article published by Andreas Backhaus, a European economist, in which he notes that there is a high incidence of COVID-19 cases in older groups in Italy, which could explain the high case fatality rate observed. However, this requires to be investigated. Why in South Korea the highest number of contagions occurs in the age groups of 20 to 29 and 50 to 59 years while in Italy it occurs among people over 70 years?



Graphs originally published here

Another example is the article by Helen Jenkins, researcher in epidemiology and infectious diseases at Boston University's School of Public Health, which reviews research that supports the effectiveness of social distancing and isolation measures to mitigate the effects of epidemics and pandemics. These investigations include the case of the Spanish flu that ravaged the world in the early twentieth century, in which there is a vision of how we can "flatten" the case curve to avoid collapsing health systems and decrease the severity of the crisis.


Image posted here

On the other hand, the wide variety of areas where AI tools have been successfully developed allows to try all kinds of applications for better monitoring, diagnosis and treating the disease. An interesting Forbes article mentions some illustrative examples, among which are:

  • Monitoring and forecasting of outbreaks.
  • Diagnosis from CT imaging via computer vision.
  • Distribution of medical supplies with drones.
  • Use of robots to disinfect and distribute food and supplies.
  • Drug development.
  • Detection of infected people.

Finally, there are efforts to call upon the data science and AI community and industry to combat he pandemic. An example of this are the challenges that Kaggle, a popular platform for collaborative and open souce development, launched to work on 10 tasks. Among them is the development of forecast models that, rather than getting right the number of cases, are sought to identify the factors that contribute to the spread of the disease. Other tasks involve using natural language processing techniques to extract useful information from more than 29,000 papers, which are impossible for anyone to read.

We are facing a disease that until today has no definite treatment and that, although it does not generate a serious condition in most cases, endangers the lives of millions of people. The most effective way to try to mitigate its effects is to reduce its spread with social distancing and isolation of the sick. This will bring severe consequences for our lives, but it is necessary to do so to avoid much greater damage.

In these serious circumstances, we must make use of the vast amount of information that we have at our disposal to know how to act, be responsible with our safety and that of our family, and hope that the efforts of thousands upon thousands of people in many different fields will pay off to better understand this threat and, hopefully, find the means to fight it effectively and prevent it in the future.

Did you enjoy this post? Read another of our posts here.

Visit our other sections