HALIFAX – Scientists at the IWK in Halifax have developed an algorithm to help doctors better pinpoint the brain’s language centres, and now the research is going international.
Research scientist Dr. Tim Bardouille, who is also a professor at Dalhousie’s School of Physiotherapy, has been working on the algorithm for the past few years, which will be used in magnetoencephalography, also known as MEG, scans.
MEG scans give doctors a 3D and 360-degree view of the brain and can offer more detail than the traditional MRI scan.
The study involves setting patients up with the MEG, showing them a series of pictures and having the patients say the name of what those pictures are.
“We can look at what’s happening in your brain in just a split second, frame by frame,” Bardouille said. “Then we can isolate the part of the activity that’s related to actually thinking of the name that you have to say.”
The algorithm is meant to streamline work for doctors.
“Getting from sitting in the scanner to a picture of your brain, there’s a lot of steps in that process. Some of them require expertise to be completed. What we’ve done is we’ve automated that entire process so it’s basically a one-click process. You get the scan. You press a button. You get the brain image out.”
The research focuses primarily on epileptic patients, although researchers are also looking at patients with aphasia and who have trouble understanding and initiating speech.
Bardouille said the algorithm is particularly important for patients with epilepsy since their brains rewire where language comes from, meaning it can sometimes be difficult to pinpoint exact locations of language centres in the brain.
He adds finding those specific centres are critical, especially if the patient needs surgery to remove the part of the brain that is generating epilepsy.
“If you’re going to do surgery on the brain, you want to be really careful about the areas that you target. Particularly to keep the areas that are most important for things like walking, talking and using your hands,” he said.
“This study is looking particularly at saving those areas that are involved in talking and understanding language.”
Currently the Wada test is used for language mapping, but it can be invasive and risky for patients. Bardouille said using the MEG scan instead is advantageous for patients because it is non-invasive, thereby more scans can be done faster, which could take patients to surgery faster.
He also said that the algorithm makes MEG scans more consistent, meaning a scan done in one part of the world will look exactly the same as a scan done on the other side of the globe.
“The MEG scans can tell you which areas are involved in producing language within a centimetre. That’s really helpful information for surgeons. But one of the problems is MEG labs all around the world will take a different approach to how they take the brain scan and turn it into a pretty picture.”
“That variability makes it difficult for us to get a large scale assessment of this technique. Having this algorithm allows us to do this in a consistent way.”
The IWK researcher has partnered with scientists in Helsinki, Finland to test the new technology in a clinical trial.
Jyrki Makela, a neurologist and senior lecturer at Helsinki University, said Helsinki has a long tradition in MEG scans.
He is hopeful for what the algorithm could add to the current process of language mapping.
“In normal subjects, it seems to be working beautifully and although we have several ways to localize speech that we use in Helsinki, we would like to improve those possibilities,” he said.
“If we can make really precise maps of individual features of speech and organization in patients, this will be tremendously useful for surgery.”
Makela expects to scan and test between 20 to 25 patients in Helsinki for the clinical trial. Roughly the same number will be tested here in Halifax.
Researchers expect results of the collaboration to be completed in one year and then, after that, they hope to expand to other centres around the world.