Dengue fever : Google Trends to track outbreaks
Published the 21.07.2017 at 18: 37
Google could well make his return in the tracking of epidemics. This time, his analysis tool queries the Internet will be used to closely monitor the dengue fever. This disease, transmitted by the mosquito, is the origin of 390 million infections per year. In the less developed countries, this remote analysis could offer a welcome solution, according to the study published in PLOS Computational Biology.
The authors of this work, and exercise at the university of Harvard (Usa), have risen the algorithm ARGO. Previously used by Google Flu Trends, which closed in 2015, it claimed to predict the evolution of the epidemics of seasonal influenza. But his lack of effectiveness has led to its fall.
The american specialists have therefore reworked the algorithm, focusing on the 10 queries most issued, and have combined the analysis of Web queries to data epidemiology government. The effectiveness of the predictive ability of ARGO has been tested in 5 countries affected by the virus of dengue fever : Mexico, Brazil, Thailand, Singapore, and Taiwan.
A system of supplementary supervision
Everywhere, except in Taiwan, the algorithm delivers a vision to be fairly accurate in real-time the evolution of the epidemic at the national level. On the island off the coast of China, on the other hand, the absence of epidemic season was clear and the low number of cases limits the impact of this monitoring.
These results confirm that the Google searches may help to better understand the progression of diseases transmitted by mosquitoes, including dengue fever, in countries where the traditional supervisory systems do not exist or function poorly.
Because, as a result of a network distended, and numerous contaminations, the signals are difficult to detect with the conventional network. The researchers hope, in the future, be able to refine the research on specific geographic zones and certain times.