An Essential Worker During COVID-19 Wasn’t Even Human

According to the U.S. Department of Homeland Security, essential workers are those who “conduct a range of operations and services that are typically essential to continue critical infrastructure operations.” These occupations, while determined by individual states, generally include healthcare workers, along with those who work in grocery stores, pharmacies, child care, energy sectors and non-profits. More than a year ago, many of us probably never really heard of the term, essential worker, yet here we are.

Alongside the heroic humans who have been stepping up to the plate regardless of their occupation, artificial intelligence (AI) has joined the front lines. It is assisting humans in their work and making the insurmountable challenge of finding a vaccine, diagnosing COVID-19, helping people get vaccinated, helping insurance claims get paid or assuaging public concerns a reality.

It’s no wonder that AI spending in the healthcare and pharmaceutical industries is expected to increase from $463 million in 2019 to more than $2 billion over the next five years, according to a recent study by ABI Research. In an industry where AI has been trying to find its place and prove it’s ready for primetime, the pandemic has helped it proven its worth.

Consider the following examples of how AI, in all its forms, has battled the pandemic.

Vaccine delivery. From Moderna, Pfizer and Johnson & Johnson to countless other vaccines still in development, the speed at which highly effective treatments have been developed is unprecedented. A process that would normally take 10 years was reduced to months. Collaboration between government and industry and accelerated approval timelines have much to do with it, but AI also played a critical role. Since science is grounded in data, deep learning algorithms have been able to examine the virus’ complex structure, and sort through thousands of components to identify those that are most likely to trigger a robust immune response. Likewise, machine learning has made it possible to create models that can learn and find patterns within available data and make inferences from previously unseen data. And, deep learning is taking that a step further, allowing features to be extracted automatically from raw data.

Innovative diagnostics. Thankfully, many people are becoming vaccinated, but COVID-19 remains a major health threat. AI is helping researchers develop ever more accurate, convenient and faster testing so that infected people can be treated and isolated as soon as possible. Machine learning and computational modeling are allowing researchers to study the ways different techniques and materials can improve the accuracy and response time of screening procedures. At the University of Oxford, for example, machine learning is helping scientists use routine healthcare data from the electronic health records of thousands of patients to predict COVID-19 in patients coming to the emergency department.

Public health management. As of mid-March, more than 448 million vaccine doses had been administered worldwide, and it’s only the tip of the iceberg. For every person vaccinated, it’s essential that the batch and lot numbers are recorded along with the name and location of every person who received one. In addition, for two-dose vaccines, there needs to be a way to easily prompt the patient to return for a second dose. It’s overwhelming to capture that information alone by administrators, so AI is helping to process that data. Much faster than would be humanly possible.

AI is also helping to fairly distribute vaccines. Last winter in Minnesota, demand for the vaccine has surpassed supply. A predictive AI algorithm is helping to identify which of the patients are in the highest risk group in order to prioritize appointments.

Claims processing. As hospitals work to treat sick patients first and foremost, the financial toll of COVID-19 is just beginning to be realized on hospitals, health systems and payers. In fact, the AHA estimates a total four month financial impact of $202.6 billion in losses for America’s hospitals and health systems. The mounting costs, claims and settlements are creating a backlog of administrative processing, yet new AI-driven document processing solutions are handling many cases so that humans can focus on the anomalies and address more strategic issues. AI is enabling faster and more accurate processing of customer data, while also helping to identify and address fraudulent claims more easily.

Addressing public concern. With strict no-visitor policies in place, hospitals are not only trying to take care of sick patients, but to also answer incoming calls about symptoms or patient updates. Chatbots are helping to answer questions, direct callers more quickly and easily than would be possible with human staff alone.

For example, Providence Saint Joseph Health turned to technology to help it more effectively manage three critical stages of care: triage, testing and treatment, relying on chatbots to assist during the triage phase. By visiting its Coronavirus Assessment Tool online, people can find out more about which symptoms might indicate the virus, and figure out if they should be seen by a health professional. This chatbot is connected to a virtual patient care visit which enables people to discuss their symptoms with a nurse practitioner. It has had overwhelming success with the public; in its first day of use alone, more than 500,000 people used the chatbot.

Finding evidence in the images. While computer vision has made its way into the medical diagnostics field, helping to detect anomalies in x-rays and other mediums, it’s being put to use in new ways. Advanced computer vision algorithms are helping to detect COVID-19 in airports, train stations and other public areas. AI-driven computer vision models are monitoring travelers to identify any suspicious breathing patterns or fever, using infrared cameras. The data is the uploaded to a server where the AI algorithm processes the data.

Another interesting way computer vision is being applied is by monitoring movement and traffic using video data. In this way, computer vision is able to read images of people in busy areas and identify locations that may not be practicing safe social distancing or mask wearing, and areas that may need better urban planning.

From deep learning models, to chatbots and computer vision, AI has become a silent partner to essential, front-line workers during COVID-19. While it has been under the very worst of circumstances, it has proven its value in supplementing the brilliant work of humans, and will continue to gain traction in telehealth, discovery and diagnostics well into the future.

Author Information

Carlos Meléndez is the chief operating officer and co-founder of the artificial intelligence and software engineering services firm, Wovenware, which helps companies achieve customized digital transformation to propel their businesses to the next level. With expertise in business strategy and software engineering, Meléndez has a strong track record bringing game-changing AI, machine learning and deep learning solutions to organizations across a variety of industries.