This text is part of the special Research section
Artificial intelligence opens the way to many medical treatments that we would not have thought possible not so long ago. Among the promising avenues: neural interfacing. At the University of Montreal, Guillaume Lajoie’s laboratory is working on tools allowing this communication between the human brain and computers.
A scientist whose work lies at the intersection of neuroscience and artificial intelligence (AI), Guillaume Lajoie is a professor in the Department of Mathematics and Statistics at the Université de Montréal, holder of a Canada-CIFAR Chair in AI and of the Canada Research Chair in Neural Computing and Interfacing. The latter focuses on the interface between AI, technology and the nervous system.
“When we talk about brain-machine interface, it’s a term that implies in particular the possibility of reading information from the brain to control external objects, explains Guillaume Lajoie. For example, we think of a quadriplegic who wants to control a cursor on a screen. On the other hand, this interfacing also concerns uses where one would like to perform a very specific neurostimulation. The objective would then be to interact with the nervous system to promote recovery after an accident, or to modulate chronic disorders to complement or replace pharmacological treatments. In general, therefore, it is a question of trying to understand how to interact with the nervous system in the same way as one writes computer programs. »
Types of interfaces
In some cases, these interfaces, which are then described as invasive, are implanted directly in the brain and record fairly precise signals there. Other, non-invasive devices, such as helmets with electrodes plugged into an individual’s head, also record brain activity, but do not have the same precision.
However, these electrodes existed long before the advent of AI. In particular, there are treatments on the market that use this type of intervention, for example deep brain stimulation technology (deep brain stimulation), where an electrode is implanted to stimulate the part of the brain responsible for the degeneration and tremors seen in Parkinson’s disease.
“Where AI becomes important is that the devices we build are more and more sophisticated, with a higher number of electrodes, and that the interaction with the brain becomes more complex. We use AI to optimize this process. This is where automated methods become very important. »
These methods are currently being developed for various health applications, with targeted medical interventions. Professor Lajoie’s laboratory is affiliated with BIOS Health, a UK company specializing in neural interfacing, which aims in particular to stimulate the vagus nerve in order to prevent cardiac arrhythmia problems.
“It is possible to treat cardiac arrhythmia by interacting with the nervous system, says Guillaume Lajoie. BIOS Health, like many others, is in the development process. The regulatory processes to bring these treatments to market are lengthy, and with good reason, but the potential is enormous. If you think about neural issues, the pharmaceutical tools that we currently have have great side effects, and the new technologies that are being developed are very promising. »
But as with the development of pharmaceutical products, the marketing of new technologies based on AI in health is highly regulated. In the context of neural interfacing, some treatments, such as deep brain stimulation, are already on the market to treat some effects of Parkinson’s disease, as mentioned, but also those of depression and epilepsy.
“Let’s be clear: there are still a lot of things that we don’t understand about the brain,” says Guillaume Lajoie. Brain-machine interfaces then become exceptional scientific research tools. Experiments performed with brain-machine interfaces are the cutting edge of important scientific questions. They have value for future clinical processes, but they also have great scientific value. »
The researcher gives as an example the interfacing used in a laboratory at the University of Washington, with which his team collaborates, to try to better understand how the brain learns. “Thus, if a subject learns a task, we have access to what is happening in the brain to accomplish it. It’s a unique tool that will allow us to ask new questions and advance our understanding of the brain, in addition to clinical treatments for pathologies,” he says.
This special content was produced by the Special Publications team of the Duty, pertaining to marketing. The drafting of Duty did not take part.