A University of Alberta research team is using artificial intelligence to make it faster and easier to diagnose and treat stroke patients, potentially saving lives and improving outcomes. The research is funded by a grant from Alberta Innovates.
The project, titled “Agile AI Pipeline for Urgent Health Conditions Requiring Diagnostic Imaging,” will receive $799,662 in funding over three years from a fund aimed at improving healthcare with AI applications. It is one of five newly funded U of A projects announced today.
“Our program provides the research funding needed to help companies and innovators identify and overcome the barriers preventing the more widespread use of artificial intelligence,” said Laura Kilcrease, CEO of Alberta Innovates. “The goal is faster, better healthcare for Albertans and developing new technologies here in Alberta that can be exported around the world.”
The goal of the project is to develop a software program that will augment information gained from brain scans taken with existing CT technology. Doctors will be able to look at the scans and determine whether a patient is having a stroke, how much brain tissue has been damaged, and whether they would benefit from reperfusion therapy with clot-dissolving drugs, a procedure called endovascular thrombectomy, which removes blood clots from blocked blood vessels, or both. If a patient requires thrombectomy, they must be transported quickly to a major center, such as the University of Alberta Hospital.
Currently, there are 3,800 ischemic strokes occurring annually in Alberta, with approximately 400 patients treated with thrombectomy. These patients typically suffer from large blood vessel blockages and severe strokes, but with prompt treatment, disabling symptoms can be reduced or even reversed.
The software provides doctors with information to ensure the right patients receive this treatment in a timely manner and prevent unnecessary and costly transfers.
“With more accurate assessments, patients should be able to receive the most appropriate treatment sooner,” said project leader Greg Kawchuk, professor of physical therapy in the Department of Rehabilitation Medicine.
“These are pretty big decisions that are made in the emergency department,” Kawchuk said, “and the new AI data on strokes provides better and deeper analysis, and this is information that physicians want to have available for the benefit of their patients.”
More efficient diagnostic and treatment decisions
Strokes are much harder to diagnose than conditions such as heart attacks because there are no definitive scans or blood tests, explains research team member Brian Buck, associate professor of neurology at the Alberta Faculty of Medicine and Dentistry and co-chair of the Alberta Health Services Working Group on Stroke. Current CT scans help rule out other potential causes of stroke symptoms, such as brain tumors or bleeding in the brain. The new system provides a map of blood flow and highlights blocked arteries, helping doctors make faster diagnoses and treatment decisions.
The project uses CT scans and AI software to help neurologists quickly determine whether a patient is having a stroke or has had a stroke in the past, which blood vessels are blocked, and whether there is brain tissue that can be saved by opening the blocked vessels with a thrombectomy procedure.
“If the answer is yes to all of those questions, the next step is often to get the patient to the University of Alberta Hospital as quickly as possible,” Buck says. “There are only two facilities in the province that can offer thrombectomy, so we're trying to improve access to thrombectomy treatment for stroke patients across the province.” The second site is in Calgary, but this particular research project is targeting communities in northern Alberta.
The team is working with Cercare Medical, a Danish medical imaging company, and five remote stroke centres in Grande Prairie, Camrose, Westlock, Fort McMurray and Cold Lake – all part of the Northern Alberta Remote Stroke System.
“Alberta has a very experienced group of stroke researchers and experts in computer science and artificial intelligence,” Buck says.
AI could help glean better information from existing medical tools: “We work within a health system with limited resources, so we want to make these decisions as efficiently as possible.”
Applying what we learn to other difficult diagnoses
With their AI and imaging projects, the team also plans to target another, less serious but more resource-intensive condition: back pain, which Kawchuk said is among the top five reasons for emergency department visits but rarely leads to hospitalization.
The AI program would read images of a patient's lower back to help doctors make faster diagnostic and treatment decisions, said Kawczuk, an adjunct professor of clinical biomechanics at the University of Southern Denmark and former Canadian Chair in Spinal Cord Function.
“We took two very different problems in the emergency department – stroke and back pain – and showcased what AI can do,” he said, adding that he hopes the software can eventually be used to diagnose other conditions that often require diagnostic imaging, such as brain injury or pulmonary embolism.
Kawchuk noted that the team plans to use an approach called federated learning, which allows the AI program to continually learn and improve across different sites without sharing specific data, ensuring patient privacy.
“The AI program will leverage existing data to make better decisions, but it will leave the data intact and come back with an upgraded structure,” he says.
Other University of Alberta projects funded by Alberta Innovates
The research is one of five University of Alberta-based projects receiving funding from Alberta Innovates. The others are:
- Bo Cao, Psychiatry, “Evaluating AI for Opioid Overdose Prediction: From Research to the Real World,” $800,000
- Luke Eckersley, Pediatrics, “SCANAID – Screening for Congenital Heart Disease with Artificial Intelligence Detection,” $798,857
- Miyoung Kim, Augustana (Science), “Developing AI to support timely and informed recommendation creation for Alberta Health Link 811,” $783,988.
- Othmar Zaianeh (Computing Sciences) and Jasmine Noble (Mood Disorders Association of Canada), “Enhancing Navigation and Access to Mental Health Services and Programs: The MIRA Artificial Intelligence Information Chatbot,” $799,447
/University of Alberta release. This material from the originating organization/author may be out of date and has been edited for clarity, style and length. Mirage.News takes no organizational stance or position and all views, positions and conclusions expressed here are those of the authors. Read the full article here.
