Diagnostics
BC network brings AI-driven ultrasound to rural sites
August 7, 2024
VANCOUVER – UBC cardiology professor Dr. Teresa Tsang (pictured) is leading a team that’s bringing advanced cardiac imaging to even the most remote parts of British Columbia. Dr. Tsang is a distinguished cardiologist specializing in ultrasound imaging of the heart, known as echocardiography, which uses sound waves to diagnose a range of other heart conditions. As the leader of the UBC-VCH AI Echo Core Lab, she is using AI to democratize and enhance access to echo imaging across B.C., particularly for those in underserved regions.
“The integration of AI in cardiac imaging has transformative potential,” says Dr. Tsang. “What used to only be possible in major healthcare centres, can now be done virtually anywhere,” she said.
The idea for the network was born during the COVID-19 pandemic. Traditionally, patients from these regions would travel to urban centres for diagnostic procedures like ultrasound imaging. However, the pandemic introduced critical risks associated with transferring patients, including the potential spread of the virus.
Dr. Tsang and her team in the UBC-VGH AI Echo Core Lab rose to the occasion. In partnership with Providence Healthcare, industry partners Change Healthcare and Clarius, and the Rural Coordination Centre of BC, and with support from Canada’s Digital Technology Supercluster, UBC and Vancouver Coastal Health Research Institute, they developed a digital network that enabled the use of point-of-care hand-held ultrasound devices by health providers in distant B.C. locations.
About the size of a smartphone or tablet, these AI-empowered devices are small but mighty. Their initial use in early COVID times enabled heart and lung images acquired from patients residing in remote and acute care areas to be transmitted to the UBC-VGH AI Echo Core Lab for rapid interpretation.
The team then went on to validate a number of AI technologies to use with the units, making it easier for even novice users, and enabling greater precision in automatic assessment of heart and valve function.
“AI is the catalyst for healthcare transformation,” says Dr. Tsang. “It’s about turning data into diagnoses, insights into interventions, and challenges into opportunities. It isn’t just innovation, it’s revolution.”
This landmark initiative, which represents a collaboration between academia, health authorities, hospitals, communities and industry, is considered the world’s first and largest publicly funded point-of-care ultrasound imaging network to support rural patients and First Nations communities and has been widely recognized as a major B.C. contribution during the pandemic. The Honourable François-Philippe Champagne, Canada’s minister of Innovation, Science and Industry, referred to it as “transformational.”
Ensuring the accuracy of AI-facilitated ultrasound imaging requires being cognizant of the fact that heart structure, function and physiology can vary across groups due to genetics, biochemistry, sex, race and other variables. Machine learning thrives on abundant data to identify patterns. Dr. Tsang’s team utilized data from hundreds of thousands of patients across B.C. to train their algorithm for accurate detection of heart function and valve disease.
“An algorithm that is developed from a more homogenous group may not work as well if applied in B.C., which has a diverse population,” says Dr. Tsang.
At Vancouver Coastal Health Research Institute, where she serves as executive director, Dr. Tsang is leading a new data and AI platform, with support from the Canada Foundation for Innovation, that will integrate patient data from health authorities in B.C. to improve the management of health conditions. Developing AI tools from these large, local data sets, representing a diverse group of patients, will help ensure that the resulting algorithms are tailored to the unique needs of B.C. communities.
“For healthcare AI to truly serve our B.C. populations, the data driving our algorithms must come from the diverse populations representative of the people that the algorithms are being used for, and that also speaks to the concept of health equity that we want to achieve for all British Columbians,” Dr. Tsang says. “Only then can we ensure equitable and effective outcomes for all.”
The UBC-VGH AI Echo Core Lab is nationally recognized for its contributions to the field and Dr. Tsang was bestowed with the Canadian Society of Echocardiography Annual Achievement Award in 2023. Dr. Tsang is committed to leveraging AI to enhance echo-imaging efficiency across the province and beyond.
“Echo-imaging is the workhorse in cardiology as it provides the diagnosis for many conditions, such as heart failure, myopathies, and various valve diseases to name a few,” she notes. “Delays in imaging will also lead to delays in treatment, including surgery if needed, further impacting patient outcomes.”
Over the years, AI applications have improved efficiencies and workflow in the echo lab. Beyond just speed, AI applications can improve image quality and enable more accurate assessment of heart structure and function. “With AI, we achieve greater precision and, in less time,” says Dr. Tsang.
However, educating physicians about AI will be vital to this evolution, as the technology continues to develop at an exponential pace. “If we don’t keep up with the rapid expansion and acceleration of knowledge, we’ll quickly fall behind,” Dr. Tsang warns.
UBC already offers continuing education courses for physicians on AI-facilitated clinical care, and Dr. Tsang is confident of the benefits. Contrary to the fears that AI might depersonalize care or replace physicians – a common fear of skeptics – AI is actually enabling physicians to use their time more efficiently to better serve their patients.
Dr. Tsang has engaged in consultations with patients and the public, discovering widespread enthusiasm for leveraging AI to enhance our healthcare system.
In April, Dr. Tsang and her team hosted a public event, funded by the CIHR Café Scientifique program, to welcome community perspectives on the development and deployment of AI tools in healthcare. Making sure that the public and patient partners have a voice in AI research is critically important so that the resulting tools meet the preferences and priorities of the communities the health system is designed to serve.
“Ultimately, our goal is the same: to create an efficient and effective system that maximizes the time doctors spend caring for patients, ensuring the highest quality healthcare outcomes for everyone in our community.”
A version of this article was originally published by The University of British Columbia Magazine and was written by Heather Gerrie.