COVID-19: How Korea is using innovative technology and AI to flatten the curve
The Republic of Korea has been successful so far in containing COVID-19 without shutting down its economy, even though schools, museums and gyms — places where large numbers of people might gather in an enclosed space — are closed.
How did they do that?
The country’s experience from the previous SARS outbreak and the more recent MERS (in 2015) has enabled it to prepare in advance to combat COVID-19.
But there’s a lot more to the success.
ITU listened to experts, live from the Republic of Korea, during Episode 1 of the ITU AI for Good webinar series, who shared their experience and response to the pandemic, and how innovative technologies have been used to help flatten the curve in the country.
“We still need to learn more about [COVID-19’s] characteristics – although we do know now that it is very contagious,” said Seon Kui Lee, Director of the Division of Risk Assessment and International Cooperation at the Republic of Korea’s Center for Disease Control and Prevention (KCDC). “Information and communication technologies are required to enhance traditional control measures and response measures, as well as for the development of innovative solutions.”
Here are some of the top reasons for success:
1) Fast-developed testing kit
An important part of the Republic of Korea’s strategy since the outbreak of the coronavirus has been widespread testing.
Thanks to artificial intelligence (AI), the development of a coronavirus testing kit happened fast in the country. Life sciences company Seegene came up with a coronavirus testing kit in under just three weeks
“To develop a test in such a short time would not have been possible without AI,” said Tai-Myoung Chung, Professor in the Department of Interaction Science at Sungkyunkwan University (SKKU).”
The testing kit, which would normally have taken two to three months to develop, was approved by the authorities within less than a week of its application, and also certified in the European Union.
Just a few weeks later, when an enormous cluster of Coronavirus cases emerged in the city of Daegu, it was ready for testing.
The number of coronavirus cases in the Republic of Korea peaked on 29 February, recording a total of 909 infections.
“Currently we have a total of 118 testing stations available nationwide, and have a testing capacity of 15,000 per day on average – with a maximum of 20,000 per day,” said Lee.
“It began with only lab testing, which later on expanded to local governments and also to the primary medical laboratories and hospitals,” said Dr Lee.
Widespread testing in the Republic of Korea is targeted mainly at the high-risk groups, i.e. those with underlying diseases, the elderly, people who share homes, or live in crowded areas, and passengers at arrival points emanating from countries with cases of the coronavirus or other infectious diseases in the past such as SARS or MERS.
2) Smart quarantine information system
Dr Lee told the webinar audience that a quarantine information system was put in place after the MERS outbreak in 2015.
“Even before this COVID-19 outbreak, inbound travellers entering the Republic of Korea have been required to be checked for fever and to also to fill out a health questionnaire,” explains Dr Lee.
Inbound passengers with symptoms or having travelled to or from a risk country are placed in quarantine.
“Information about the inbound traveller from the Ministry of Justice, the Ministry of Foreign Affairs, airline companies and major telephone telecommunication companies are collected by KCDC’s quarantine information system,” says Dr Lee.
This enables frontline health workers to have a full record of the patient’s history of movements to help them quickly identify and isolate or treat the suspected coronavirus patient in a timely manner.
In-bound travellers are required to download a self-health check mobile app to their smartphone and submit their health conditions on this app during their 14-day incubation period, explains Lee.
With the collaboration of telecom companies, they also receive texts messages and receive guidance on how to report any coronavirus symptoms they might have developed while in quarantine.
3) Mobile phone technology data for contact tracing
As well as interviewing, officials are using location data from mobile phones, credit-card transaction records and CCTV footage to trace and test people who might have recently come into contact with an infected person.
In many places, detailed maps are published showing precise movements of infected people, encouraging others who thought they might have been in contact with an infected person to seek out testing.
4) AI for improving diagnosis efficiency and patient classification
Lee explained that as part of the risk-mitigation strategy, KCDC established a system which categorized confirmed cases into four categories: mild, moderate, severe and very severe.
“Each category receives a different treatment and is admitted to a different facility according to the severity of the case,” said Dr Lee.
Professor Chung gave examples of AI-based tools that are being used in the country to enable the quick diagnosis and classification of patients.
For example, VUNO‘s Chest X-Ray AI Image Support Decision Tool – an algorithm for identifying abnormal findings on chest x-rays – classifies intensive care patients by using X-ray images and can examine the lung within just three seconds.
JLK inspection, though numerous studies have developed an all-in-one medical platform called AiHub for disease diagnosis, which, it says, uses world class AI and big data technology from various imaging devices. It can examine lung disease within seconds using an AI technique that is being used in hospitals.
The company has also produced an AI-based, hand-held chest X-ray camera which can scan the chest in just three seconds and give a heatmap visualization of abnormal lesion.
5) Mobile apps for information sharing
Many mobile apps have been very quickly developed in the country since the outbreak of the virus.
They have proved useful for information sharing to advise and inform the public. “This would be a huge task for already stretched health-care staff and volunteers,” claimed Professor Chung, saying that they can be developed quickly within a few days to a week.
For example, one mobile app has been developed to direct those who develop symptoms to the nearest available testing station. Another can indicate the nearest point of purchase for available masks.
A public chat robot using AI techniques is being used to inform on ways of responding to coronavirus, and another AI-based voice robot automatically calls people who need attention, explained Professor Chung.
6) Daegu: Making use of a smart city hub
Daegu City, where the majority of the Republic of Korea’s cases of COVID-19 have occurred so far, is currently in the process of being transformed into a smart city (set to be completed in 2021).
The epidemiological investigation during the outbreak was able to use the data hub of the smart city, particularly for tracing patient routes, which Chung says is “critical to developing a new medicine as quickly as possible.”
Find out more about how the Republic of Korea used ICTs and AI to flatten the curve on the coronavirus by listening to the recording of Episode No. 1 of the AI for Good Webinar Series.
The AI for Good Webinar Series is a free, live series of talks, interviews and panels, featuring inter-disciplinary experts whose ideas, insights and solutions can help humanity leverage AI for good.
Don’t miss Episode No. 2: Covid-19: Using mobile phone & AI for contact tracing while respecting privacy on Friday, 3 April 2020.
Register now, for the AI for Good Global Summit (21-25 September 2020).