News

» Go to news main

Dal professor receives grant for researching improved MRI tech

Posted by Terry Murray-Arnold on November 23, 2023 in News, Research, School of Health Sciences
Dr. Steven Beyea is a professor of diagnostic radiology with cross appointment to the School of Health Sciences.
Dr. Steven Beyea is a professor of diagnostic radiology with cross appointment to the School of Health Sciences.

A Dalhousie professor has been awarded a grant of just under $280,000 for his latest work on MRI technologies.

Dr. Steven Beyea has been researching diagnostic imaging technologies for over 25 years. He currently works as a professor of diagnostic radiology with cross appointment to the School of Health Sciences.

With the new funding, courtesy of the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery program, Dr. Beyea’s lab will receive $279,167 over a five-year period. By the end of this grant, Dr. Beyea will have held NSERC Discovery funding for 25 consecutive years.

His latest project is focusing on improving the quality of diagnosis that patients receive by optimizing magnetic resonance imaging (MRI) technologies. As better imaging technologies emerge, Dr. Beyea wants to ensure that the right quality image is used for the right job.

“There’s a tendency to think that a diagnostic image from an improved technology is better healthcare, but those two don’t always equate,” Dr. Beyea said in an interview.

In Dr. Beyea’s research, he has conducted trials, showing radiologists images of varying quality from MRI scans. Participants were asked to rank how confident they were in making a diagnosis based on each image.

The results found that, in some cases, a better image from an MRI can improve the confidence of a radiologist’s diagnosis. In other cases, however, it changes nothing, and the radiologist reaches the diagnosis with the same level of confidence as with a more rapidly acquired less detailed image.

“There are methods that exist for objectively, quantitatively saying this image is better quality than this one … but the issue is that they were developed for natural photographs,” Dr. Beyea said. “If you and I were to look at some of these MRI images, we likely would not do a great job predicting what a radiologist thinks of the clinical quality of that image.”

The goal, then, is to increase image quality only when necessary. If a lower quality image is all a clinician needs to make an accurate diagnosis, it means a quicker procedure for a patient which can lead to higher efficiency for our healthcare system and greater comfort for patients.

All of the money from the grant will fund graduate and undergraduate students working in Dr. Beyea’s lab.

“A focus for me is always having, what I would call, three legs on the stool; we can’t develop healthcare technologies without healthcare,” Dr. Beyea said. “We can’t do it without the discovery of new ideas, hence involving graduate students who are, from an academic perspective, driving new ideas. … and in virtually every project I do involve industry partners such that if we find something we can actually get it out.”

Artificial intelligence is also playing a role in the research, using “deep learning” techniques to improve image quality at a faster rate than current technology, providing a clearer and less distorted image.

A.I. is also helping to determine which cases require a high quality image, and which cases only require a less detailed but potentially faster image for diagnosis. Using deep learning training, the A.I. is provided data from trained radiologists, with the ultimate goal being that the A.I. will determine what quality is required for a given procedure, saving time and resources.

Beyond the technical work, Dr. Beyea says ensuring industry partners are involved once enough progress has been made is an essential part of the process.

“I want this to be something … that’s not just a paper, but it’s out there impacting patients,” he said.

With an ongoing shortage of healthcare staff in the Maritimes, Dr. Beyea is confident that there is high potential for this technology and others like it. He hopes that his research can contribute to better, quicker, and more accurate healthcare in the province.

Nova Scotia in particular has a shared health database system between all provincial hospitals, meaning data can be pooled across individual hospitals, especially with the introduction of the One Patient One Record system.

“We have opportunities here where I think we can innovate, given the nature of our healthcare system,” Dr. Beyea said. “We can demonstrate leadership in how we can and should do these things.”