COVID-19 has been rapidly spreading across the world reaching over one million confirmed cases and more than 60,000 deaths to date since the first reports of the virus in December 2019. A sudden surge in cases of a virus with an unknown cure can wreak havoc on any healthcare system. Many countries are struggling to cope with the outbreak as their health centers and clinics are overwhelmed by the influx of patients.
With most businesses being forced to close and billions of people practicing social distancing and lockdowns, these circumstances can put a huge strain on society. Brave healthcare professionals all over the world have been working around the clock to contain and treat the novel coronavirus, establishing efforts that will hopefully restore the calm.
To deal with the public health crisis, some institutions have been exploring Artificial Intelligence options to help professionals manage patient flow and hospital capacity. Advances in AI applications such as machine learning, natural language processing, and computer vision can be useful in helping researchers, doctors, and nurses to diagnose, predict, and treat COVID-19, as well as reduce the risk of misinformation.
Virtual assistants and chatbots
As hospitals and their call lines get swamped by the rising number of COVID-19 cases, patients calling in may encounter longer than normal wait times that can extend to hours. In some cases, call centers can experience a 50% increase in call volume. As a result, institutions have had to devise new ways of operating and keeping people informed on the virus.
Choosing to adapt to the circumstances, some companies and hospitals are turning to chatbots and conversational virtual assistants to help manage these spikes in call traffic. The World Health Organization launched a messaging service, which has the potential to reach two billion people, dedicated to providing the latest news and information on COVID-19.
These AI-based answering systems have the ability to detect which consultations can be handled by virtual assistants and which require a person-to-person interaction. Due to the rise in telecommuters and people living in lockdowns, many people have turned to these services as a reliable source of medical information.
Virtual assistants can be trained to answer common questions related to COVID-19, provide information and clear guidelines, as well as advise callers on whether they need to self-isolate at home or, in the most extreme cases, seek professional help at a nearby hospital.
Health organizations can integrate voice assistants, such as Vozy’s Lili, into their processes to help manage patient flow and alleviate pressures caused by high demand in consultations. Automating responses to patients calling in to reschedule appointments, verify lab results, or review a patient’s healthcare plan can help organizations double their efforts during trying times.
Tests that receive a delayed diagnosis are less likely to be accurate, according to experts in the industry. This means that hundreds of thousands of people can end up believing that they have tested negative for coronavirus only to be asymptomatic carriers unknowingly propagating the outbreak. Therefore, quicker diagnoses using machine learning could ultimately save lives. Already, medical professionals are using AI models to perform tasks such as reading CT lung scans.
These AI solutions are not meant to do the job on their own but can serve as a powerful tool to help a skilled radiologist. Some of these technologies have the potential to determine if 200 of 500 images require closer inspection by a radiologist, significantly saving doctors time that could instead be used to tend to other patients.
Some countries have started to monitor highly transited places like airports, hospitals, or restaurants using thermal imaging to automatically detect individuals with fever. Since the absence of a fever does not guarantee that a person is not carrying the virus, this AI application can be helpful for screening but cannot be relied on for an actual diagnosis. In some cases, technology can recognize faces and determine whether a person is wearing a face mask or not. If a person’s temperature is too high, it could result in them being denied access to a building.
Hospitals and other health institutions can use AI mapping tools to identify high-risk communities during the spread of COVID-19. These tools take into consideration the location of nearby hospitals, food sources, as well as transportation, which could reveal some of the social determinants of health that increase the risk of contraction. Having access to these findings could help determine where to focus efforts and prioritize care for the most vulnerable patients.
Although some of these AI solutions are still in the early stages of development and application in combating the pandemic, the preliminary results of introducing these technologies in healthcare are encouraging. Combined with the expertise and bravery of healthcare professionals, AI solutions have the potential to add great value to and relieve the overwhelmed healthcare industry during the COVID-19 outbreak.