COVID-19 continues to wreak havoc on society. Fortunately, artificial intelligence is offering up solutions to stem the spread and more.
With artificial intelligence seemingly “just around the corner” – from self-driving cars to instant machine translation – AI has more potential uses than the average user might guess. And that includes healthcare; amidst the current COVID-19 pandemic, AI is being deployed in a number of surprising ways. How is AI being used? Further, what has the COVID-19 pandemic made plain about our digital world? Read on…
1) Hospitals and healthcare facilities need to be cyber-protected
Hospitals and healthcare providers are clearly critical to our economy and lives. Yet, medical facilities have been bombarded with cyber-attacks such as phishing emails, ransomware and malicious malware. For instance, ransomware took down the website of the Champaign-Urbana Public Health District in Illinois. Because medical care has become more networked and interconnected via computers and devices, hospitals are logical targets for hackers. Enterprise networks and medical devices such as diagnostic imaging systems, infusion pumps, medical apps, lasers, electrocardiographs, pumps, monitors and ventilators are also hacker targets.
2) Data needs to be secured
According to Forbes, integral factors to mitigation strategies have been predictive analytics and forecasting surrounding the spread, infection rates and the lethality of COVID-19. Data is more important than ever for our economic preparedness and economic prosperity and we need to be able to organize, manage and analyze growing amounts of data. Yet, this is not a small challenge. Businesses are facing growing risks in data loss, with the U.S. experiencing 1,244 data breaches last year with 446.5 million exposed records.
3) Digital transformation is accelerating
As of 2020, most of our critical infrastructures act in a digital environment, such as the healthcare, communications, transportation, energy and financial industries. This dynamic digital environment is connected to the IoT which is comprised of billions of devices and trillions of sensors. As we will discuss later in this article, artificial intelligence and digital sensors used to track COVID-19 have been useful for drug discovery and applications for therapeutics. As digital transformation continues to be underway, components will rely more on cybersecurity, big data analytics and cognitive tools.
Companies are finding a multitude of uses for learning more about coronavirus, how to stem its spread, and ultimately, a cure. For instance…
AI Can Drugs that Target the Virus
According to ZDNet, there are a multitude of research projects that are using AI to identify drugs that can be repurposed to take on coronavirus. Companies want to be able to identify which drugs can disrupt the way COVID-19 work by studying the molecular setup of existing drugs with AI.
For instance, London-based drug-discovery company BenevolentAI began working on finding a cure for coronavirus in late January. The company’s has an AI-powered knowledge graph that can analyze large volumes of biomedical research and scientific literature to find links between the biological and genetic properties of diseases, as well as the composition and action of drugs.
According to the company, due to the amount of data that is being produced about COVID-19, it was able to adapt the knowledge graph to take into account the latest information about the virus itself, as well as the kinds of concepts that are more important in biology.
Gaining Insights into the Structure of COVID-19
The AI arm of Alphabet, Google’s parent company, DeepMind, is currently using data on genomes to predict organisms’ protein structure, which can possibly shed light on which drugs could work against the coronavirus. They have released a deep-learning library called AlphaFold, which uses neural networks to predict how the proteins that make up an organism, based on their genome, curve or crinkle. Why? Protein structures determine the shape of receptors in an organism’s cells and once you know what shape the receptor is, it becomes possible to tell which drugs could bind to them and disrupt vital processes within the cells. For instance, when it comes to COVID-19, it disrupts how it binds to human cells or slows the rate it reproduces at.
Detecting the spread of new diseases and the outbreak
When coronavirus was still localized to Wuhan, AI systems were thought to be among the first to detect the outbreak. Some have thought that HealthMap, which is affiliated with the Boston Children’s Hospital and is an AI-driven system, picked up the growing cluster of unexplained pneumonia cases before human researchers. Though it only ranked the outbreak’s seriousness as “medium,” it was able to identify the earliest signs of the outbreak by mining in Chinese language and local news media like WeChat and Weibo. This data has been made public and available to scientists and researchers searching for links between the disease and certain populations, including containment measures. This data is then combined with data on human movements to see how population mobility and control measures affected the spread of the virus in China. To this day, HealthMap continues to track the spread of coronavirus throughout the outbreak by visualizing its spread across the world by time and location.
Spotting signs of a COVID-19 infection in medical images
DarwinAI, a Canadian startup, has developed a neural network that can screen X-rays for signs of the COVID-19 infection. Analyzing chest x-rays could give hospitals an alternative to testing via swabs when they don’t have enough staff or testing kits to process all their patients quickly.
COVID-Net was released by DarwinAI as an open-source system. More datasets of x-rays were contributed to train the system, which has now learned from more than 17,000 images. In the meantime, researchers from Indonesia, Turkey, India and other countries are all now working on COVID-19.
Monitoring how mental health is affected
Using AI-driven text analysis, Johannes Eichstaedt, an assistant professor in Stanford University’s department of psychology, queried over two million tweets hashtagged with COVID-related terms during February and March. He then combined it with other datasets on relevant factors including the number of cases, deaths, demographics and more, to illuminate the virus’ effects on mental health.
The analysis showed that much of the COVID-19-related chat in urban areas was centered on adapting to living with, and preventing the spread of, the infection. Rural areas discussed adapting far less, which Eichstaedt attributed to the relative prevalence of the disease in urban areas compared to rural, meaning those in the country have had less exposure to the disease and its consequences. There are also differences in how the young and old are discussing COVID-19, with younger counties discussing subjects like hand washing more.
Forecasting how coronavirus cases and deaths will spread across cities – and why
Kaggle, a Google-owned machine-learning community, is giving a number of COVID-19-related challenges to its members, including forecasting the number of cases and fatalities by city as a way of identifying exactly why some places are hit worse than others. Currently, the community is working on a dataset of infections in 163 countries from two months of this year to develop models and interrogate the data for factors that predict spread.
Many of the community’s models have been producing feature-importance plots to reveal which elements could be contributing to the differences in cases and fatalities. The research has found that it’s the policies and cultural norms in countries that are the reasons that transmission rates are different in different countries.
Coronavirus isn’t going away any time soon. Fortunately, we have tools at our disposal to identify outbreaks, slow its spread, and more.
Is your business looking for solutions amid the pandemic? Stefanini is here to help. Check out the many ways we are responding to the crisis here.