1 April 2020

COVID-19 & Artificial Intelligence

By Dr Joel Than Chia Ming

Artificial Intelligence (AI) has been a growing area of research and study in the world. From predicting stock markets and weather storms to broad medical applications such as breast cancer and lung disease detection, AI has shown it has a place in the world today.

At present, the world is gripped with uncertainty and fear due to the COVID-19 disease. Governments around the world are putting travel restrictions and stricter policies in place to curb the rising number of infections and deaths. Scientists as scrambling to find cures and hastening for human testing.

The first area where AI is starting to play an important role during this uncertain time is in the stopping and curbing of fake news or information. The year of 2020 has seen not just the widest spread of virus but also the widest spread of misinformation. This is largely due to the increase of accessibility to internet and social media presence.

Fake news about magical cures such as salt water and raw onions are spreading like wild fire and preying on the gullible. Giant social media platforms such as Facebook, Twitter and Reddit have recently issued statements that they are fighting the war of COVID-19 linked misinformation. Facebook AI, for example, has already successfully removed millions of fake user accounts and is filtering posts.

Current work of AI uses deep learning, a form of AI that checks the stance of posts. The stance mentioned here relates to the process of determination of whether a certain document agrees, disagrees or takes no stance on a specific claim. AI can filter millions of documents or postings at a single time as compared to a single person reading one posting at a time.

In China where the virus began and has the most cases and fatalities, AI is beginning to be deployed for disease detection. Paired with a powerful computer hardware, an AI system is put in place to detect COVID-19 through chest-scans of Computed Tomography (CT) scans. So far, the success detection rate is claimed to be 80% but with further improvement and modification, the accuracy could potentially be higher.

The AI system has been developed not just to be executed in China but also for international market with an English interface. Not only does it give a prediction of disease presence, it also manages to direct doctors to specific lung regions that require special attention by circling them in different colours. This particular implementation also lists possible next steps the doctor can take. These next step suggestions are based on the experiences and lesson that are inputted from doctors who have had real world experience with COVID-19 patients.

Elsewhere, companies are researching to understand COVID-19 better with AI. A better understanding among researchers will allow them to make better decisions on treatment and containment. DeepMind Technologies, a company based in UK, has shared its AI software’s finding that shows the details and structures of six proteins that are linked to COVID-19. This sharing was done in the hope that it will help scientists who are currently developing the vaccine.

The amount of research for COVID-19 has also increased tremendously due to vested interest of every community. This is seen in the higher number of publications. However, compiling and consolidating these research works to understandable milestones and for indications of next steps is an important endeavour. This is when AI comes into the picture. By using Natural Language Processing (NLP), it can automatically group and compile research that is published to show the trends and most read publications on COVID-19. 

The White House has just initiated a project called the COVID-19 Open Research Dataset, also known as CORD-19, which is a call to action for AI researchers around the world to attempt such an endeavour. The goal of this project is to aid researchers to better analyse and understand a growing set of scientific publications regarding coronavirus including COVID-19. This initiative powered by AI can help avoid information overload and also pave the way for researchers to find new information on how to address the current virus outbreak.

AI has shown great potential and undoubtedly will play a greater role in time to come. Stephen Hawking once said that “AI could possibly be credited as being the next best thing to happen to humans”. Thus, let us be witnesses to this AI rise in our world today. For those with keen interest in learning about AI, Swinburne Sarawak offers a three-year Bachelor of Computer Science degree with a major in Artificial Intelligence.

Dr Joel Than Chia Ming is a lecturer at the School of Information and Communication Technologies, Faculty of Engineering, Computing and Science, Swinburne University of Technology Sarawak Campus. He can be reached via email at jcmthan@swinburne.edu.my.