How AI And Data Science Are Helping To Understand And Tackle COVID-19

The COVID-19 pandemic has changed our daily lives in ways we could not have imagined just over a month ago. For many of us, our lives now happen within four walls. Around the world, restaurants, shops and theatres have been closed. Millions of people are in an increasingly precarious financial situation. As of the 8^th April 2020, almost 1.5 million people have tested positive for COVID-19 worldwide, and over 80,000 have died from it. The race is on to control the spread, find a vaccine and still help to support the millions with other health issues, all while supporting the economy.

Helping to understand COVID-19

The AI company Black Swan Data has turned its skills to help clinicians, researchers and the medical community in their work on analysing COVID-19. They’re used to making sense of billions of data points, often in public conversations. Black Swan’s sister charity White Swan, which applies the same data tools with a mission to improve the health of society, has been working on a search engine for academic papers on coronavirus.

White Swan is working on developing an NLP (neuro-linguistic programming) tool to search 44 000 research papers related to coronavirus. The aim is that any clinician or researcher can search for answers without having to read through thousands of papers. A coronavirus research team at Imperial College London and a doctor who is dealing with hospitalised patients are helping White Swan to verify results and refine the search model. The model is looking to help answer key questions, including what is known about the transmission, incubation and environmental stability coronavirus? What do we know about COVID-19 risk factors? And what do we know about non-pharmaceutical interventions?

At a time when primary or secondary care workers are not treating many cases, White Swan is also looking at innovative ways to analyse COVID-19. They are investigating whether public social data can generate insight on diagnosis, treatment, prognosis and the effectiveness of PPE (Personal Protective Equipment). The lack of taste and smell has been one such symptom which has gained as much traction through social conversations as clinical evidence.

In line with this research, they also answered to an AI challenge with The White House, AI2, CZI, MSR, Georgetown & NIH posted on Kaggle, a machine learning data science community. The challenge asks for AI experts to develop text and data mining tools that can help the medical community develop answers to high priority scientific questions. They’re now voluntarily working on this to see if they can help build a search engine specifically for academic research purposes.

Supporting more patients

The ongoing challenge for health systems around the world is not only how to cope with tackling COVID-19 but also continue to provide healthcare to other patients when resources are stretched.

White Swan has already been worked with the NHS to predict A&E attendance and accelerate diagnosis is continuing that work with an app called Million Minds. This app analyses thousands, if not millions, of public data and conversations on illnesses and helps to link symptoms. This work helps to speed up the patient pathway and get a diagnosis faster.

Nightingale is another project White Swan is working on to empower patients and optimise self-management. The project started with patients of ankylosing spondylitis, a rheumatic disease. The app uses data from wearable devices to better understand what impacts the severity of different symptoms.

Linked trends and fads

The core work of Black Swan Data is to analyse attitudes, behaviours and cultural shifts that are happening to help businesses and governments make informed decisions. Phil Norminton, Head of Insights, talks about the high tension in conversations around concerns for family, financial security and health that rose in March. According to Black Swan data on the U.K., conversations around a better immune system have grown by 560% in March compared February, before the COVID-19 crisis. The first strong peak in conversations around panic buying was on 2nd March.

Comfort snacks are clearly on the rise alongside immunity-boosting foods. In the U.S., supplements such as vitamin c, elderberry, turmeric, ginger, ashwagandha and sarsaparilla are growing in conversation. At the same time, conversations on cinnamon rolls are also up by 66%. In the U.K., there’s a preference for chocolate digestives which have been talked about 147% more in March compared to February.

On the other hand, new ways of connecting and learning are also shooting up. There has been an 87 000% increase in conversations about the ‘Houseparty’ app and a 6000% increase in ‘Google Classrooms’ conversation. With people becoming more self-reliant or simply stuck at home, conversations on “How to guides” are also on the rise by 33%. For the moment though, Phil Norminton says it’s too early to know which will become long-lasting trends or if they’re short-term fads.

The application of AI and Data Science

Data science has a wealth of applications to societies issues and is helping to tackle this global crisis. Steve King, the CEO of Black Swan, loves the fact that you can look at weird and wonderful data and make it useful. This global pandemic has undoubtedly shown how useful that analysis and those tools can be. Data Science and AI are playing an important role to accelerate understanding and solutions around the world, especially when time is of the essence.