Seemingly overnight, communities and businesses around the world have been forced to change countless behaviors and systems in the pursuit of a “new normal.” Although many of the headlines have pivoted to business and financial struggles, there remains an urgent need to protect and ensure human health. This involves many elements, from extending the capacity of healthcare systems and developing effective tests to establishing tracking systems and enabling researchers to make headway on a vaccine. These activities are all focused on human needs and conditions, yet technology can play a key role in moving each one forward.
Artificial intelligence (AI) can be particularly important, with a diversity of functions that enables the healthcare community to deploy it in many different ways. AI clearly has a role in the creation of COVID-19 chatbots, and it can be used to scour volumes of papers and studies to accelerate coronavirus research. Data-driven approaches might also supply information that helps implement mitigation and suppression strategies to combat the spread of COVID-19.
Yet AI can play an even more-practical role as overburdened healthcare systems try to cope with the deluge of people who want to be tested. And it starts at home.
The global nature of this pandemic has resulted in considerable data about COVID-19 symptoms and the medical resources available to individuals. If AI were leveraged to conduct cognitive triage at home, enabling patients to enter their symptoms and answer a series of questions in conversational form, this “symptom bot” could provide answers that may reduce citizens’ anxiety and minimize the number of people going to hospital unnecessarily.
The vast majority of COVID-19 cases can be managed from home with guidance from a physician or medical expert, and less than 5% of cases require ventilation. As a result, many doctors and healthcare associations have asked people to stay home to avoid potentially infecting other citizens and healthcare workers. An AI tool could enable this.
Individuals could also use AI to send automatic inquiries about home testing or to receive personalized guidance about the best possible treatments based on specific data (age, existing conditions, etc.). After a virtual visit, patients could then receive customized care instructions and automated follow-up inquiries from their provider’s office. This entire interaction could be handled more efficiently than an on-premise visit, would reduce the risk of human-to-human transmission, and could alleviate the burden on hospitals that are already struggling with capacity issues.
Such a solution wouldn’t come without challenge. For instance, access to data points is crucial to leveraging big-data algorithms and AI, data quality and standardization are vital to delivering consistent answers, and each engagement must be handled with the utmost privacy to meet regulatory and medical-record requirements.
Fortunately, such obstacles are common – and overcome – in many other AI deployments. As the healthcare community leans on technology for many facets of the COVID-19 response, AI may be a powerful tool in their kit. Not only useful for chatbots and research, AI can also help address the tangible challenge of alleviating strain on an overburdened healthcare system.