AI in Outbreak Management: How AI is helping the fight against COVID-19
With millions of recorded infections worldwide, the need for effective measures to halt the pandemic is more pressing than ever, but how is AI being used to combat the spread of COVID-19?
Countries are using AI-based systems in a number of ways to help manage and control the spread of the coronavirus. These applications include research for the development of a vaccine (by speeding up genome sequencing), assisting healthcare professionals in frontline services, supporting research and promoting knowledge worldwide, forecasting the evolution of outbreaks, and support for measures that restrict the movement of populations. However, like any other technology, AI has its own limitations and challenges. Dr. Nadine Nehme, Chief Science Officer at Medicus AI, helps us learn more.
Research into AI comes from both the public and the private sector. How has the world’s scientific community responded to the crisis so far?
The genuine threat of a pandemic has shown us what humanity is capable of and that humanity will rise to its challenges. The sheer amount of scientific literature available reflects the eagerness of researchers to deal with this major health crisis.
In the weeks following the discovery of the novel coronavirus, over 2000 research papers were submitted. Subsequently, Microsoft Research, the National Library of Medicine, and the Allen Institute for AI (AI2) collected and prepared more than 29 000 documents at an unprecedented pace. While such a huge collection of scientific literature is clearly impressive, it also represents a real challenge for anyone hoping to actually use the information that has been collated.
There’s also the question of “fake news”. The influx of uncontrolled information leads to misinformation, but there are organizations working to combat this. Inserm has launched a website that locates misinformation about COVID-19, while Unicef developed a chatbot to help people quickly access accurate information from a reliable source.
Other than publishing scientific papers, what sort of responses have been made?
Some organizations have started more dynamic initiatives: Kaggle, a Google subsidiary, created a series of challenges to find the answers to ten key questions about COVID-19. These questions aimed to forge a greater understanding of topics such as risk factors, non-drug treatments, the genetic properties of the virus, and vaccine development efforts.
It’s been almost half a year since the first COVID-19 cases were reported, but there’s still no effective treatment available. Does the involvement of AI mean that a vaccine will be found faster?
Hopefully, yes. The role of AI in medical research is to aid researchers in finding a vaccine faster. There are certain technologies that can assist in the rapid development of a protein vaccine that is testable on humans, such as the one recently released by American startup Moderna.
For example, protein-folding algorithms such as LinearFold (by Baidu, Oregon State University, and the University of Rochester) and AlphaFold (by DeepMind) can significantly increase the speed with which protein structures are solved.
These types of epidemiological, bioinformatical, and molecular modeling technologies require significant processing power to realize. This is where technology giants such as IBM, Amazon, Google, and Microsoft are assisting.
Are there ethical concerns regarding the use of AI to fight COVID-19?
Yes, as is the case with all uses of AI. With the stakes so high, there is a balance to be struck between protecting human rights and confronting a public health crisis. Some societies, such as Singapore, accepted the emergency social restrictions more readily than others, like the US.
The crisis has also given rise to increased state control: mass-surveillance policies in China equipped law enforcement agencies with “smart” helmets capable of recognizing individuals if they are wearing masks (with 95% accuracy) and flagging individuals with high body temperature.
According to the European GDPR (Article 9), personal data collection and analysis are allowed, as long as the intended use has a clear and specific public health aim. However, the efficient use of AI is hampered both by a lack of data and by too much data. Using AI ethically means maintaining a careful balance between the two, which will lead in turn to an equilibrium between data privacy and public health.
How effective has AI been so far?
AI is proving to be an important tool in building a coordinated response to the pandemic. However, there are several challenges. The effectiveness of emergency measures that use technological solutions, including AI, should be reassessed at the end of the crisis.
While AI has many uses, from predicting the spread of the novel coronavirus to powering robots that can replace humans in hospital wards, systems must be in place to implement these solutions effectively. The structural issues encountered by healthcare systems in the current crisis are not down to the technological solutions themselves, but due to the inadequate organization of the health services. AI alone cannot compensate for these structural difficulties.
The AI-based solution developed by Medicus AI, CoVive, is a comprehensive COVID-19 probability assessment, monitoring, and analytics user app and enterprise platform.
An appetite for a little more…
How we built it: Covive
When we saw all the various solutions and apps being launched for COVID-19 in early March, we got to work. We quickly realized that there was no app that combined all the possible use cases we had in mind to serve people’s needs simply and easily. That is when we got to work building an app that did everything we wanted COVID-19 app to do and launched CoVive as a global initiative.
