Disruptive technology is first to put harmful plaques under an ‘AI microscope’, assigning risk scores to patients to help prevent strokes from happening

Montreal, QC— June 2, 2022— Karina Gasbarrino is on a mission to improve the prediction and prevention of strokes — the second-leading cause of death and third-leading cause of disability worldwide — so fewer families have to endure the sudden loss of a loved one. She’s achieving her goal with a first-of-its-kind tool she developed that uses cutting-edge artificial intelligence (AI) and image analysis technologies to do just that.

Her breakthrough work, a cloud-based system called SonoPlaque™ that harnesses AI to more accurately examine harmful fatty deposits in the arteries of the neck, has earned Gasbarrino a prestigious award from Mitacs, a national innovation organization that fosters growth by solving business challenges with research solutions from academic institutions.

In recognition of her efforts to advance this first-of-its-kind platform through her digital health startup, PLAKK Inc., Gasbarrino — who completed her PhD in Experimental Medicine at McGill University and now serves as PLAKK co-founder and COO — will be presented the Mitacs Social Entrepreneur Award on June 2 at a ceremony in Montreal.

Gasbarrino’s journey to prevent strokes started more than a decade ago when she lost her grandfather to a sudden stroke. Since then, she has devoted her academic career to his memory, helping to advance early detection and diagnosis of unstable atherosclerotic plaques in carotid arteries in the neck. Rupturing of these plaques is the main cause of strokes.

Now her company’s platform is set to revolutionize stroke prevention by allowing clinicians to accurately identify and characterize high-risk plaques using ultrasound images, applying deep learning to understand what those plaques are made of and measure key parameters that indicate high risk of stroke. Patients are assigned a stroke risk score to help determine if preventive surgery or treatment is required, similar to the way a high PSA level helps determine if intervention is required to treat prostate cancer.

“Coming out of my PhD, I knew I wanted to have a major impact on patients’ lives and I felt the way I could do that was to start thinking about a tool that could be easily integrated into medical clinics,” said Gasbarrino. “Right now, stroke events are happening anytime, anywhere, and we just don’t know when or where that might be. We know plaques cause the majority of strokes, but we’re not screening for them.”

The problem with current diagnostic methods for stroke prevention is that treating physicians only consider one parameter: the level of artery blockage as shown on a patient’s neck ultrasound and reported by a radiologist. Yet, research shows one parameter is insufficient to tell the exact risk of stroke and therefore people are missed.

SonoPlaque™ goes a step further by providing more information about the plaques themselves — such as the amount of inflammation, fat or calcification present — to help to quantify the risk. Clinicians upload their patient’s ultrasound images to the cloud and the platform automatically assigns a stroke risk score, using the deep learning models developed by PLAKK to quickly analyse plaque parameters. If the stroke risk score is high, the plaques can be surgically removed or treated with confidence, proactively preventing stroke.

The company’s key differentiator from competing approaches that are also trying to automate the process of characterizing plaques in the neck, is that it uses low-cost, safe ultrasound images. Other emerging technologies use CT or MRI imaging which expose patients to radiation, are more costly or not as widely available as ultrasound. “We’re also the only approach to apply precision medicine using AI and routine ultrasound images, and to assign an individual risk score,” said Gasbarrino.

Plakk is currently working on two versions of its technology, one that will serve as a decision support tool for radiologists reading the ultrasound images, and one that will be tailored for family physicians to use, particularly in rural and remote communities. This year, it expects to get an early prototype evaluated by radiologists in the field as it works towards achieving Health Canada and FDA approvals.

Gasbarrino is one of five winners of the Mitacs Entrepreneur Award who are being recognized for their efforts to turn their research into an innovative business that impacts the lives of Canadians.

“Mitacs allowed me to transform a scientific passion into a disruptive technology by giving me the opportunity to spend countless hours working on the foundation for our innovation and validating its clinical need,” Gasbarrino said. “Over the past two years, Mitacs has also provided support to other members of PLAKK’s R&D team, helping to accelerate the development process for our innovative platform.”

“Mitacs is committed to helping up-and-coming innovators through their entrepreneurial journey, and we’re extremely proud of the remarkable accomplishments of each of this year’s award winners,” said Mitacs CEO John Hepburn, adding that 20 percent of Mitacs interns successfully turn their innovations into startups. “The success of our country’s entrepreneurs in commercializing ground-breaking innovations not only goes a long way in boosting Canada’s economic future, but also helps put Canada on the map as a research and innovation leader.”

About Mitacs

Mitacs is a not-for-profit organization that fosters growth and innovation in Canada by solving business challenges with research solutions from academic institutions. It is funded by the Government of Canada and the Government of Ontario, along with the Government of Alberta, the Government of British Columbia, Research Manitoba, the Government of New Brunswick, the Government of Newfoundland and Labrador, the Government of Nova Scotia, Innovation PEI, the Government of Quebec, the Government of Saskatchewan and the Government of Yukon.

For information about Mitacs and its programs, visit mitacs.ca/newsroom.