Artificial Intelligence To Prevent A Health Crisis
The World Health Organization notified the world on January 9, 2020, that a pneumonic outbreak, reported in Wuhan province, China, was developing. Still, the US Center for Disease Control was three days ahead of them, revealing the worrying development since January 6 that year.
Surprisingly, beating both institutions by a week, the first to warn of the potential danger that COVID-19 presented to the world was not an international institution or a select group of researchers, but BlueDot, an artificial intelligence. Using an algorithm that allows it to explore news in multiple languages dealing with disease trends in plants and animals worldwide, the Canadian platform advised its clients to avoid danger zones, such as the Chinese province where the virus originated.
When responding to pandemics, speed is crucial, and the program was developed looking for a way to analyze astronomical amounts of information in times that would be impossible for a human being, yielding results that could be used to prevent massive contagion.
And while other AI programs for the same purpose have failed in the past, BlueDot was also correct in predicting the location of the first Zika virus outbreak in the United States. This new success gives hope to the rest of the current efforts in this technology field that seeks, from various approaches, to put artificial intelligence at the service of humanity as assistant epidemiologists.
Artificial Intelligence Is The Result Of International Cooperation
Combining algorithms from the companies Huawei (China) and Siemens (Germany), the Radvid-19 tool analyzes X-ray readings and chest CT scans looking for irregularities in lung images. The technology is currently implemented in 43 Brazilian hospitals, unifying the analyzes of all the doctors who use it in that country, creating a network of measurements that allow more accurate data to be obtained the more it is used. Although it must be clarified that the evaluation tool cannot replace virus tests, it has been a “substantial help in the diagnosis of cases in doubt”, according to radiologist Arthur Lobo, who works in northern Brazil.
Almost on the other side of the world and from a completely different approach, Luxembourg saw the birth of the CDCVA (COVID-19 Detection by Cough and Voice Analysis) project, which proposes an innovative detection based on the voice and cough patterns of infected people. Its creator points out that “respiratory conditions caused by COVID-19 can make patients’ voices different, creating identifiable voice signatures that can be recognized using our system.”
In Wuhan, where the global outbreak began, applications were deployed that enabled detection of fevers among large crowds of people, implementing variants of this technology at bus and train stations across the country. This allowed the Chinese authorities to measure the population’s temperature without contacting them, determining the appropriate containment and sanitation actions from a distance.
Another critical use that has helped countries like Germany and South Korea, exemplary in their handling of the virus, was the implementation of artificial intelligence similar to BlueDot that, by collecting and analyzing data, increased the analytics collected and the number of determining tests carried out.
The access and analysis of this data involve privacy issues that are currently very worrying. For this reason, the most effective solution to detect infections is also the most controversial.
Contact tracing applications raise the possibility of using the geolocation technologies present in all smartphones to detect possible cases of contagion and notify the user of each phone of contact with infected people. Although this has proven to be the most effective way to reduce contagion, in China, there are objections against this technology that centres around the importance of privacy.
The reality is that tracking technologies should not be viewed only in a negative light but with great caution and regulation, reducing access to information only for humanitarian aid purposes, complemented by cybersecurity barriers that maintain the privacy of its users.
Also Read: The Future Of Artificial Intelligence
