Since the announcement of COVID-19 as international pandemic on March 11th, the social and economic life in all countries around the world was affected severely.
In response to the rapidly spreading infection, countries have set strict limitations to movement, travel and economic activities. People are accordingly searching for all possible tools to end this situation or at least limit its impacts as much as possible.
Artificial intelligence (AI) and machine learning proved to be a powerful and useful tool that can provide unique and creative solutions. This is especially important in monitoring and forecasting COVID-19 trends.
Leaders from Google, Nuance and the Health Data Analytics Institute have agreed in a recent virtual conference that AI became an essential and major tool to help patients and healthcare providers.
AI will also play an important role in resuming the economic activities after the end of the pandemic and determining the people who are most in need for support. The points below explain how AI helps in monitoring and controlling COVID-19 trends.
The use of AI also helps in keeping people away from places with increased probability of infection transmission. Healthcare providers are increasingly using telecommunication to provide diagnosis, medical care and support to patients which is currently known as telehealth.
Artificial intelligence started playing a major role in telehealth to save time for healthcare providers to focus on providing care to hospitalized patients.
Developers at Nuance Communications and Health Data Analytics Institute have developed virtual heathcare assistants that are capable of understanding natural language. They use chatbots to answer common questions from concerned people.
Additionally, computers equipped with artificial intelligence are gathering information from patients, analyzing the data and organizing it into more useful forms to be used by healthcare providers and decision makers.
Data collected from these chatbots and computers is also used to draw an interactive map for hotspots and numbers of reported patients. They can also detect possible changes in COVID-19 trends, which may raise alerts regarding possible mutations in the virus.
The use of AI offers the advantage of making informative decisions based on thoroughly analyzed statistics. This function is an essential tool during this unprecedented situation where healthcare resources and logistics can be significantly limited.
AI can be used to determine the most appropriate way to use resources as PPE, ventilators or treatments with limited availability such as plasma taken from recovered patients.
There is currently shortage in these resources and making the decision regarding their use is an extremely stressful and painful decision for healthcare providers since it comes to the lives of people they care for.
Managing resources using AI can also be beneficial to alleviate the financial impact of the pandemic on the healthcare system. Much of the revenue for hospitals comes from elective procedures that are currently cancelled.
Accordingly, hospitals are trying to implement business models that can withstand the financial distress. They are also trying to automate the documentation, coding and managing claims to maintain their financial income which is crucial to avoid the collapse of the healthcare system.
The use of AI can have more profound role in the scientific research community regarding COVID-19 trends and nature. Synchronizing the work of brilliant minds and powerful research labs around the world is urgently needed to reach a satisfactory treatment or vaccine to the infection.
There are around 200 papers published daily about COVID-19. It is obvious that this huge amount of data cannot be efficiently traced by researchers working in the field.
In response to this issue, the White House Office announced a call to action in partnership with leading partners from the academia and industry as IBM, NASA, Microsoft and Massachusetts Institute of Technology.
The aim of the call is to develop a tool that uses the supercomputing capabilities available in the US to efficiently fight the pandemic. In response to this call, researchers at Lawrence Berkeley National Laboratory developed a prototype of a text-mining tool that was called COVIDScholar.
This tool uses natural language processing and machine learning to gather research related to COVID-19. The objective of developing the tool is not only finding the published articles but also gathering information, analysis and finding unnoticed COVID-19 trends, connections and insights.
Developers of COVIDScholar say that the tool may be an additional step towards “Automated Science”. COVIDScholar was able to gather more than 61,000 papers about COVID-19 in one month. Developers of this search engine say that it provides more relevant results.
They also aim at providing deeper insights and conclusions such as finding genetic links between the disease and human phenotypes. Reaching such result would require analysis of huge amount of data that is not normally possible using conventional methods.
On the larger scale, countries are using AI to monitor the effectiveness of the implemented measures and to forecast future COVID-19 trends. The data is also used to determine the suitable time for reopening, taking into consideration the health and economic impact on people.
For example, the Organization for Economic Co-operation and Development (OECD) developed AI tools to monitor the numbers, develop trajectories related to the disease in addition to predicting the outcome of current social distancing regulations. The tools are also available for public use as an educational version.
Additionally, the organization provides an updated collection for COVID-19 related AI initiatives. The ultimate objectives of the OECD.AI is to provide evidence based analysis to set better policies and regulations.