Unlocking the injured brain

Photo by JD Howell


On Feb. 19, 2019, Toronto Maple Leaf centre Nazem Kadri took a series of big hits in a game against the St. Louis Blues – and missed the next game because of a concussion.

While concussions in professional sports make occasional headlines, they happen with surprising frequency to everyday people – every 11 seconds in the U.S. and Canada, according to John Connolly, a professor in the department of Linguistics and Languages at Mac and the Senator William McMaster Chair in the Cognitive Neuroscience of Language.

And it’s not just sports that cause head injury– concussions can happen to people in automobile collisions, who fall down the stairs, take a spill in the kitchen, or tumble off their bikes.

Right now, if you get a bang on the head and suspect you might have a concussion, the standard clinical approach is to conduct an interview to collect information about symptoms and perhaps run a few behavioural tests to help determine how seriously you’re injured.

That’s a problem, Connolly says.

That’s because behavioural assessments can be highly subjective, relying on judgments from both patient and clinician that essentially amount to educated guesses. If a brain scan, most commonly CT, is conducted it likely won’t show any concussion-related damage, and can’t determinine what the cognitive effects of the concussion might be.

“There’s no shortage of behaviourally based tests, and behavioural judgments by the person themselves, or the clinician, as to the severity of the concussion,” explains Connolly. “But how do you objectively answer a question like, ‘How bad are your symptoms?’”

No objective measure has been readily available to assess concussion and traumatic brain injury patients – until now.

Connolly, working with a number of colleagues over the years, has spent the majority of his career figuring out how to get the injured brain to “speak” when regular speech isn’t possible or isn’t effective – either because a patient is in a coma, is otherwise non-verbal, or, in the case of concussion, simply isn’t aware of the specific damage that’s been done.

To do that, he uses electroencephalography, or EEG, to record the electrical signals the brain produces. Using machine learning techniques – including algorithms developed with Mac colleagues – Connolly and his team analyze the results to pinpoint specific anomalies that are generated by injured brains: confirming the presence and severity of the cognitive consequences and, most importantly and uniquely, providing a road map for treatment.

Concussion patients, for example, are given tasks to perform while their brain activity is measured with an EEG. Each task tests cognitive functions like working memory, automatic attention, reactive attention, concentration, language comprehension or executive function. Completing the tasks triggers an event-related potential, or ERP – a specific type of brain activity that is then compared to healthy controls to determine the type and extent of any injury. The concussion results can all be evaluated without the need of a baseline, with changes over time during rehabilitation being the key feature of study.

This kind of objective assessment of real-time brain function is unique in the world right now – so to fill the gap between his university research and the needs of clinicians, Connolly has commercialized his work, starting a company called VoxNeuro that will officially launch in June. By providing the only objective measures of real-time cognitive activity, VoxNeuro’s assessments are able to identify specific areas of reduced cognitive function and then inform and accelerate cognitive rehabilitation as well as track recovery during therapeutic intervention.

Working with partners at the McMaster Industry Liaison Office (MILO) and Mac’s start-up accelerator, The Forge, VoxNeuro is currently a team of nine and is building connections both across the country and internationally – for example, they’ll be hosting a booth at the annual meeting of the International Brain Injury Association in March 2019.

The research activity upon which VoxNeuro rests is continuing to evolve – with, for example, new work by two anesthesiologists at Hamilton Health Sciences who are looking at the potential for the technology to be used to measure cognitive activity during surgery as well as perioperative cognitive dysfunction. Similarly, Connolly and his research colleagues were recently awarded a Collaborative Health Research Project (CHRP) grant for ongoing research into coma. However, for now VoxNeuro’s primary concern is its success in the concussion realm.

“What we provide is, right now, unique in the clinical concussion sphere – objective, reliable, replicable concussion results that are directly from the brain,” says Connolly. “This is good for patients, but it’s also good for clinicians who, at the moment, are doing the best they can in a vacuum. They get standard hospital tests, which they don’t find very useful, and behavioural measures, which aren’t reliable. What we do is different.”

While VoxNeuro is focusing on concussion assessments, it’s not only concussion patients who benefit from these types of tests. Connolly also uses EEG to assess the brain function of patients who are in a coma as well as those diagnosed as in a vegetative state (now called unresponsive wakefulness syndrome). While it had long been assumed that these patients were too brain damaged to benefit from rehabilitative interventions, this isn’t actually true. Connolly points out that some coma patients cycle between a deep unconscious state and a state of consciousness without awareness. They remain in a coma by all outward signs but their brains are responding to sounds presented to them in ways that reflect a type of automatic attention.  The on-again off-again cycling of this brain activity indicates that a patient could benefit from treatment even before they fully “wake up.” This cycling also holds prognostic value in predicting emergence from coma.

“What this really does is give medical teams a head start to determine what’s going on with a patient who is in a coma or in a state of unresponsive wakefulness,” says Connolly. “People know what to do to treat these patients, but they aren’t able to tell who might benefit from treatment. That’s what we can do.”

Having begun his work with unresponsive wakefulness syndrome patients 25 years ago, Connolly says he’s seeing a definite shift in how his work is regarded, both by clinicians and by potential business contacts.

“We were the first group in the world to publish this kind of work [on unresponsive wakefulness syndrome],” he says. “We had trouble publishing it because it was questioning the accuracy of traditional diagnostic methods – methods that are now known to be incorrect up to 43 per cent of the time.”

The methods for this work are patented and another patent is pending. Now Connolly and colleagues have developed their methods to include concussion, with important work involving retired professional football players and a longitudinal study of concussions in children and adolescents.

“It’s a different world now in terms of the reception of our work,” Connolly notes. “It may be immodest to say, but the work we’re doing is advancing the field.”

Related Stories

CHANNELS