- Health Is Wealth
- Posts
- Coronavirus: AI predicted epidemic
Coronavirus: AI predicted epidemic
[ad_1]
Technology at the service of health. Artificial intelligence predicted the coronavirus epidemic 10 days before the WHO talked about it, and researchers are creating a map to track its spread in real time.
The fields of application of artificial intelligence are almost endless. No wonder therefore that AI models exist to anticipate the appearance and spread of epidemics, such as that of the coronavirus in China. One of these AIs, called "BlueDot" and developed by a Canadian start-up, had thus detected favorable signs for the appearance of this virus in Wuhan, capital of Hubei province, as of December 31. Either 10 days before the World Health Organization (WHO) or the media start talking about it.
BlueDot had even anticipated the emergence of the first infectious foci as well as the first displacements of the virus, taking for example into account the numerous plane trips planned during this period between Wuhan and Bangkok, but also Taipei and Seoul.
AI already used in the past
Founded by a former doctor in 2014, BlueDot had already distinguished itself in previous epidemics (notably that of the Zika virus in 2016). Its algorithms are able to capture and cross countless so-called "weak" signals observed on the Web, to calculate a risk and generate forecasts. They will therefore draw on data publicly available on the Internet (requests made on search engines by regions, messages published on health forums, etc.). Much more accurate than those of the late Google Flu Trends, BlueDot reports are sent to public health officials in several countries, as well as to airlines or hospitals.
We do not know, however, what is the margin of error of this AI, which could explain why its first alerts were not necessarily taken seriously.
A real-time map to track coronavirus
Another example of technology applied to tracking the spread of viruses: American researchers from Johns Hopkins University in Baltimore have put online a global map to follow the evolution of the epidemic in real time. This map is based on official WHO data and data from various official sources (several centers for disease control and prevention and hospitals).
The goal, of course, is not to panic the populations, but to help the authorities and the citizens to better anticipate the displacements of the virus, to fight as effectively as possible against its propagation. Proven cases are counted, as are the number of healings and deaths. The path of the virus can also be analyzed in a second step, to improve the effectiveness of the management of such epidemics in the future.
Currently, the spread of the coronavirus (4,475 confirmed cases) affects 13 countries outside of China, including France, the United States, Germany and Japan.
[ad_2]