“Without clear planning and implementation of the steps that I and other experts have outlined, 2020 will be darkest winter in modern history.” — Dr. Rick Bright, former director of the Biomedical Advanced Research and Development Authority.
This is dark news indeed, and it comes on the cusp of certain countries and parts of the U.S. beginning to open for business as usual. Personally, I’ve seen Facebook friends posting family pics of a fun-looking and seemingly normal vacation at a Texas beach. Sunbathers and kids sharing a crowded beach with no masks or social distancing anywhere in sight. Images that would normally make me smile caused me to cringed in fear at the sight and wonder how my friend and so many others could be so cavalier about this virus.
Bad times are coming, but I needed to find a bright spot to focus on today. I take heart in the fact that the greatest scientists and virology experts are hard at work looking for answers. And in addition to that, technology experts are doing their part for the assist.
Throughout history, you can see the world has suffered through many pandemics. The most deadly one to happen recently was the Spanish flu of 1918 and 1919. Infecting one-third of the world’s population with 20 to 50 million deaths, the Spanish flu was especially virulent.
Looking further back in time brings us to the bubonic plague of the 14th century, Europe’s flu pandemic in the late 19th century, and an outbreak of cholera shortly before the Spanish flu pandemic. And now, once again, the world is suffering through another global disaster.
The global pandemics in the past swept through the world rapidly, ending only when a large number of people had died and most of the living had recovered from the disease and developed an immunity.
The goal of social distancing today is to buy us time to find a vaccine and drug treatments while avoiding huge numbers of losses. But the world has grown weary of lockdown, and time is running out.
In what ways can AI help?
AI is already being used in various ways to fight the pandemic.
In predicting the pandemic before it began.
BluDot, a Canadian startup firm, warned its clients days before the World Health Organization (WHO) or the Center for Disease Control and Prevention (CDC) issued statements about the virus in Wuhan, China. How did they do it? They used algorithms to search for patterns in their data. Finding a pattern emerging in Wuhan, they quickly warned clients not to travel to the area.
And beyond that, the company correctly identified the next countries where Covid-19 was going to crop up next by analyzing flight patterns.
Predicting when patients require hospitalization and ICU services.
AI is predicting when patients that have tested positive for Covid-19 will have severe cases and need to be hospitalized. As it turns out, AI has been more accurate than human judgment alone in helping hospitals to prioritize the use of their limited resources.
One example comes from Denmark, where computer scientists have begun using AI to predict when patients may require the use of a ventilator. The system takes data from patients and uses it to find out which symptoms and traits in patients have required the use of ventilators.
Using that information, it predicts which patients in the future may end up with more severe symptoms. By assessing the number of ventilators needed, doctors and hospitals can work to obtain an adequate amount of equipment before the onslaught occurs. And patients will be able to get the treatment they need.
Disseminating large volumes of research materials.
New research is being done every day on Covid-19. The challenge becomes collecting and analyzing all the information to find connections and insights that can help guide scientists looking for answers.
A tool, created by a team of scientists at Lawrence Berkeley National Laboratory is scanning tens of thousands of research papers in the hunt for insight or a connection that may not be apparent without the help of machine learning.
Using AI combined with network medicine to predict drug treatments.
Network medicine is using a complex set of interactions at the molecular level to view a disease. Combining the use of AI with network medicine has helped scientists to identify drugs at a much faster rate than without the use of technology. Drug discovery is typically very slow, but the need for a quick treatment has prompted scientists to use AI to scan through massive amounts of data to get quick answers.
One result of this comes from the London startup, BenevolentAI. Their AI software has helped to identify a drug that may dampen the effects of Covid-19. Clinical trials on the drug are set to begin in the US this June.
The challenge is in the data
It’s more than having a brilliant algorithm for machine learning. AI software can’t be developed or tested without large amounts of data to come up with valid predictions. We need high-quality data from accurate sources.
To develop an AI algorithm, testing and validation need to be done before the machine learning is ever put into actual use. For this to work, accurate data is needed to test the system and analyze the results.
Real-world data is always more complex than any training model. To accurately build and test an AI algorithm the data must meet the following requirements.
- Be free of human bias.
- Contain a large enough sample to be representative of the population.
- Include a complete set of data, not just the data that works during testing.
The fight against Covid-19 needs funding
In the race to stop this disease, the world needs to work together like never before. Funding is needed for research and testing. Some organizations have already committed to helping.
- The European Union has raised over $8 billion for vaccine research and medical treatments at the Coronavirus Global Response Summit.
- The Bill and Melinda Gates Foundation has committed to funding over $250 million for a data-led approach to fighting Covid-19.
- Rockefeller Foundation and Mastercard Center for Inclusive Growth have committed themselves to identify ways in which data can be used to fight Covid-19.
Can the world work together?
Although arguably the most overused phrase these days, we are all in this together, takes on new meaning when thinking of the collaboration occurring right now from people all around the world. The only way to fight this disease is by using everything we’ve got and we need to do it together.
Now the question remains, will governments rise to the occasion and use every tool humans have designed to stop needless suffering and death? I imagine that some will and some will not. This will play out all over the world like one giant virology experiment and only time will tell the winners and the losers.