An interdisciplinary team of graduate students and researchers at the University of Waterloo is developing an AI-powered monitoring system to prepare for future pandemics.
GoodLabs Studio, a business with close links to UW, is harnessing the power of machine learning and artificial intelligence (AI) to provide real-time, data-driven insights to healthcare authorities for decision-making in the case of a future pandemic. They are creating a system that can detect increases in atypical symptoms as they believe this is a key step in pandemic prevention.
The Department of National Defence (DND) of Canada solicited submissions for inventions that would improve the response to future pandemics in early 2021. Co-founded by University of Waterloo alumni Thomas Lo, GoodLabs leaped at the chance and has now obtained two DND grants to build the Syndrome Anomaly Detection System (SADS).
SADS uses comprehensive disease monitoring to discover trends of abnormal disease across areas and alert healthcare decision-makers and policymakers. The SADS app anonymously records symptoms stated during the patient-doctor discussions. The data is then collected and processed using deep language machine learning in order to discover increases in unusual symptoms in the community and assess the danger of spread.
The team used natural language processing (NLP) AI technology within the app to ensure privacy. Only the necessary facts — symptoms, age, gender and location — are gathered and aggregated, and the patient’s personal information is secured.
“We fundamentally believe there is an unbounded opportunity for positive impact,” said Lo. “We aim to deploy the Syndrome Anomaly Detection System in hospital triages, clinics, telehealth and eHealth forums — a system that can provide authorized government and health entities early warning of the next pandemic and its spread pattern.”
Machine learning analytics are used in the SADS backend platform to code symptoms according to the International Classification of Diseases (ICD-10) and score how unusual they are. SADS creates a statistical depiction of how a new illness could spread in an area by recording atypical symptoms over time. The technology creates an alert with critical information about a possible epidemic and sends it to health and government officials in real-time.
“We’ve learned from COVID-19 just how fast-moving pandemics are and therefore how valuable reliable data in real-time is for understanding risk,” said Dr. Jean-Paul Lam, special advisor and team lead for AI pandemic discovery on the project from the Department of Economics.
In a simulation of a COVID-19 epidemic in 2020 in multiple Canadian cities, the researchers discovered that Toronto had a detectable outbreak a week before the city proclaimed a lockdown. If SADS had been available, the simulation shows that a more proactive approach would have been taken at the time.
SADS might be used locally, nationally and worldwide, with the ability to aggregate health data gathered all over the world – at least, that is the goal.
Medical practitioners, AI researchers, software developers and an economist make up the interdisciplinary SADS project team. Lam has previously worked with Lo on a blockchain and cryptocurrency project and is a specialist in econometrics and machine learning in the banking sector. “Economists have a tendency to get engaged in topics they shouldn’t,” he joked.
Lam’s AI group has affiliations with the Faculties of Mathematics, Science and Arts at the University of Waterloo. Lo and Riyaz Somani, his GoodLabs co-founder, are proud Waterloo graduates who frequently work with UW.
“Thomas is a very natural leader. He brings this passion, not only for what he does but [also] to bring together people from very different disciplines and make them work well together while keeping an eye on the ball,” Lam said about Lo.
“It’s a super interesting project to work on,” he said. “It was really about ‘How can we help?’ I think that was great motivation for us. Once it’s deployed, we believe the SADS technology will make a difference.”
