The use of artificial intelligence (AI) at the CHU of Quebec-University Laval can support more patients for magnetic resonance imaging tests (MRI) since last January.
As part of a technological project, AI is deployed in the reconstruction of the images used in radiology and radio-oncology.
“It is really an important revolution that changes the game,” said Éric Poulin, Ph. D., medical physicist at the CHU.
Each MRI device with this technology can accommodate approximately 25 more patients per week, since the exam time is reduced thanks to the reconstruction of images. Until now, three on nine MRI devices have used AI, but deployment is underway in the various hospitals in the CHU. It is estimated that all the devices will have AI by the end of next year.
“This represents about 15% more examinations per week,” added Mr. Poulin.
Since the start of the project, nearly 2,000 additional MRI has been carried out.
From left to right: Doctor Marc-André Hamel, Radiologist at the Quebec Child-Jesus Hospital, Vanessa Blouin, Deputy Director of the Medical Directorate of Hospital Services of the Quebec-University Laval, and Éric Poulin, Ph. D., medical physicist at the CHU de Québec.
Photo DIANE TREMBLAY
“This is the first phase of implantation. There will be other new products that will follow with the next generations. With what we have as a technology today, 25 or 26 more patients per week, it’s very good, “he added.
The AI accelerates the acquisition of images which are then examined by radiologists to be interpreted. When it happens that a patient moves in the tunnel, the quality of the images can be altered, which the AI also corrects.
Prostate from an MRI test. On the left, standard protocol. On the right, image rebuilt by AI.
Photo Diane Tremblay
For several years, CHU MRI devices have been welcoming patients 24 hours a day, 7 days a week. In this context, how to create more capacity?
“It is a project like this one that improves access. We are not in the facilitation of reading, on the medical component side. We are really focused on promoting access to the population by increasing our time slots to our existing equipment. Thanks to this project, our examination times are shorter, “said Vanessa Blouin, assistant director of the medical management of hospital services of the Quebec-University Laval CHU.
Wanting is one thing. Power is another. The challenge was not to convince the managers to embark, recalls Mr. Poulin. It was rather to make sure that technology was timely before initiating its deployment, which required several validation works and clinical studies using standard protocols, all with the approval of Health Canada.
“This is a first Quebecer and even a Canadian first with such a broad deployment. We are not just there with clinical deployment, but it is also a specific project with which we will measure the impact on patient access, “said Poulin.
The Quebec CHU Foundation has granted $ 464,400 to finance part of the AI IRM introduction project. Existing MRI devices are compatible with AI. Implementation costs are more linked to the acquisition of licenses and the replacement of computers when necessary. The CHU performs more than 50,000 exams in medical imaging per year, while 10,602 patients are pending.