Algorithms that detect the brain’s hidden consciousness can help doctors predict a patient’s recovery from a head injury

Families of people in an unresponsive state due to severe brain injury face the permanent and impossible decision of whether or not to continue life-sustaining care. To make matters even more complicated, there is no definitive way to know what a patient’s chances of recovery are, and traditional bedside exams can miss important signs.

Early on, neurologists found the first whispers of a promising way to find out what was really going on in a patient’s brain, through brain activity scans. The idea was to analyze these measurements for signs of secret consciousness, which are signals that patients cannot otherwise express with movement or speech but are present in neuroimaging.

Researchers from Columbia University and New York Presbyterian Hospital have taken a big step in this research direction. A new study published Wednesday in Lancet Neurology shows that patients with signs of secret consciousness had shorter recovery times and better outcomes than those without.

Brian Edlowa Massachusetts General Hospital and Harvard Medical School neurologist who was not involved in the study called the new research “groundbreaking.”

“What families want to know is, ‘Will my loved one be able to regain a quality of life that is meaningful and acceptable to them, and how long will it take for my loved one to get there?’ “, he told The Daily Beast. “This study goes a long way to answering both of these questions.”

The study followed 193 patients with acute brain injury during their first year of recovery, registering at three, six and 12 months after injury. The researchers applied a machine learning model to electroencephalograms (EEG, which record brain waves) of unresponsive patients to detect brain activity in response to verbal commands and determine whether a patient is displaying covert consciousness.

Of patients who recovered within a year, 64% who showed a secret conscience did so within three months, compared to 29% of patients without a secret conscience.

John Claassendirector of critical care neurology at Columbia University and lead author of the study, told The Daily Beast that the work builds on a paper 2019 by his team which found an association between secret awareness and positive outcomes, but was too small to establish a causal link. This new study replicates their findings in a larger cohort and was able to show a marked benefit when controlling for other factors that influence recovery, such as age.

“As an academic community, we need to replicate and validate these findings in larger multicenter studies, but, in my view, it’s not too early to start thinking about how these findings might affect our approach to prognosis. in ICU,” Edlow said.

“What families want to know is, “Will my loved one be able to regain a quality of life that is meaningful and acceptable to them, and how long will it take for my loved one to get there?” »”

— Brian Edlow, Massachusetts General Hospital

Claassen said independent validation will help make a compelling case for changing the current standard of care and that “this document will fuel the discussion.” In the meantime, he is already devising ways to scale technology fairly in healthcare centers. Publishing the code for his team’s machine learning algorithm is one step, he said, but the centers will need to work together to collect and analyze brain activity data.

“What needs to happen is that you can do it in any place, in any hospital, in any rehabilitation place, ideally,” Claassen said.

Sharon D. Cole