The Effects of AI in COVID-19 Detection

Canada developed a platform using AI called “BlueDot”. BlueDot detected COVID-19 back on 31st December 2019, which is a week earlier than the Worldwide Health Organisation (WHO) - since the spread is exponential, being a week earlier is actually very significant. It was also able to predict the spread of the virus as it scattered through news, reports and global airline ticketing data. This allowed it to detect which group of people were infected and where they were going - BlueDot predicted COVID would spread from Wuhan to Bangkok and Japan. Furthermore, using BlueDot, researchers were able to list the top 20 cities where passengers from Wuhan would arrive and warned these cities about the possible pandemic. BlueDot was built with a $9 million dollar investment which is relatively low-cost for an AI tool that can out predict humans in spotting infectious diseases.

There is another platform called “HealthMap”, based in Boston Children’s hospital, uses AI for spotting infectious diseases and it even detected COVID-19 on 30th December 2019, a day earlier than BlueDot. However, it categorised it as a low level of significance, meaning a human was still needed to interpret the data. 30 minutes after HealthMap sent the alert, a scientist at the Program for Monitoring Emerging Diseases (ProMED) also issued an alert. In this scenario, AI was faster than a human by only 30 minutes. This shows currently, AI isn’t always that much faster than humans and the model wasn’t ready for COVID-19 detection as it couldn’t identify it as a serious threat straight away like BlueDot.

References:

Naudé, W. (2020). Artificial Intelligence Against Covid-19: An Early Review. Retrieved 3 June 2020, from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3568314