Making the world more pandemic proof - one patient consultation at a time
Soaring infection rates, rising medical costs, overburdened healthcare systems, and millions of lives affected globally. What started as a black swan event back at the start of 2020 continues to be the most pressing concern for governments, businesses, and patients worldwide. Patients worried about infecting their loved ones would call their doctors to check their symptoms and receive medical advice on treatment and quarantine protocols. To avoid the risk of infection, when a patient reports COVID-19 symptoms, they must be remotely triaged and prioritized. If they are under-triaged, their and others' health will be at risk; if they are over-triaged, valuable resources will be used ineffectively.
As of today, over 1.5 million deaths have been reported on the official COVID-19 case tracker Worldometer.
Faced with the unsettling facts, the Danish government sought to go beyond confinements and leaned on the research community to find a new type of technology to bet on, to combat the pandemic. Together with Innovation Fund Denmark, they chose to provide funding to the artificial intelligence company Corti.
Corti's approach to tackling the pandemic was simple. They believed if we don't keep our medics safe, we can't ensure patients' safety. To help healthcare departments protect staff and resources, Corti had learned to automatically pre-diagnose COVID-19 symptoms by just listening to a doctor-patient phone call. This led to patients getting the right treatments faster, keeping medics safe, and lowering the resource waste of protective personal equipment, a crucial tool to reduce the medical staff infection rate.
In March 2020, a consortium of thought-leaders in the prehospital sector was assembled around Corti to help drive the project forward. Corti's artificial intelligence Audia was then trained on 86,252 patient consultation recordings. Each consultation was cross-referenced with EHR data, showing who was an actual positive COVID-19 case. This allowed Audia to recognize key indicators of COVID-19 and recommend appropriate actions, such as reminding first responders to wear personal protective equipment (PPE) and thus limit personnel infection. By training on this large bulk of consultations, Audia learned to discriminate between a low -and a high-risk COVID-19 case, enabling medics to get an instant second opinion on how to support the individual patient.
Before long, Corti's artificial intelligence was implemented in chatbots, phone systems, and self-triaging solutions across the globe with the principal value proposition of relieving the immense strain under which the global healthcare sector operates.
We saw it as our duty to apply our technical expertise to support the healthcare professionals battling the rising number of infections. The partnership with our consortium of world-class partners enabled us to reallocate our resources and take a deep dive into developing a novel digital solution to detect and treat COVID-19 faster and more accurately. Lars Maaløe, CTO at Corti
As Audia started to have thousands of patient interactions each week, patterns started emerging. Audia was able to create COVID-19 heatmaps, enabling decision-makers to predict where COVID-19 hotspots would appear, which went above and beyond what any other COVID-19 solution was capable of doing at that point.
The data from the heatmaps sadly showed that people who had the least were hit the hardest.
People on the streets and in homeless shelters were among the first to get infected, and here the spread of the virus was unstoppable. Many already suffered from other chronic diseases and were highly at risk of getting infected, leading to a high mortality rate and the virus spreading like wildfire. Audia was able to help pick up early signs of the virus spreading, enabling local healthcare and emergency service departments to diverge resources and focus on areas with a higher risk.
As artificial intelligence solutions like Corti's have proven to do more than ever before to support healthcare professionals in a time of need, global pandemic response strategies will have changed forever. The digitization of health has long been coined as a disruptive force, but we have lacked real-world evidence of deep-tech solutions that actually had a real impact on patients' health. But the willingness to experiment with new approaches during the COVID-19 pandemic has allowed us to leapfrog new developments in artificial intelligence, medicine, and patient engagement, which will have effects far beyond the present crisis.
"One of the most intriguing learnings from COVID-19 seen from Corti's perspective was, that Audia, our artificial intelligence platform wasn't just able to learn to detect COVID-19, but we saw its capacity for learning from a patient in Florida and leveraging that learning in a faraway city like Copenhagen. This encourages me a lot when I imagine what's to come for AI in healthcare since we might still find a way to make healthcare expertise more accessible if we can transfer insights from one part of the world to another." Lars Maaløe, CTO at Corti
As the pandemic continues to evolve throughout the second wave, governments feel more prepared to support medical professionals with disruptive tools to meet the challenges more quickly and accurately than before and enable them to save even more lives across the world.