MONTREAL, Sept. 29, 2026 – How can a person regain consciousness after a severe brain injury? Why do some treatments produce remarkable effects in certain individuals while remaining ineffective in others? A new study led by Guillaume Dumas, researcher at the Centre de recherche Azrieli du CHU Sainte-Justine, sheds new light on these questions through a computational model capable of simulating the functioning of damaged brain networks.
Developed in collaboration with Italian psychiatrist Francesco Achilli of the University of Ferrara, whose advanced training fellowship in neuroinformatics within Guillaume Dumas’ laboratory inspired this study, and with neurobiologist Jean-Pierre Changeux of the Institut Pasteur in Paris, this model reproduces different types of brain injuries associated with disorders of consciousness. It enables researchers to observe how neural circuits respond to various disruptions in their activity and to explore brain mechanisms that remain impossible to observe directly in humans.
Understanding Disorders of Consciousness
Disorders of consciousness encompass a range of neurological conditions that impair a person's ability to perceive their environment, interact with it, or respond to stimuli. Some individuals appear awake but show no signs of reacting to those around them, while others exhibit fluctuating or extremely limited signs of awareness, such as in a coma.
These disorders can occur following a traumatic brain injury, stroke, or certain neurological disorders. Their diversity poses a major challenge for clinical teams, as each person presents a unique pattern of brain damage and recovery potential.
Recreating an Injured Brain... Inside a Computer
To better understand the mechanisms underlying these complex conditions, the researchers developed a computational model of the brain built from biologically realistic neurons. Unlike the artificial neural networks used in AI, whose "neurons" are highly simplified mathematical units, each neuron in the model reproduces the electrical behavior of a biological neuron: it fires electrical impulses, called action potentials or "spikes", and communicates with other neurons through synaptic connections. As in the human brain, most of these neurons are excitatory, meaning they promote the transmission of nerve signals, while a smaller proportion are inhibitory and help regulate this activity.
The team then simulated lesions of varying sizes and locations, affecting excitatory neurons, inhibitory neurons, or both simultaneously. By analyzing their effects on network activity, the researchers were able to explore different hypotheses and gain a deeper understanding of how neural circuits respond to various physiological disruptions. This approach provides a controlled environment for studying the consequences of brain injuries and the mechanisms that may support the recovery of brain function.
"We do not have direct access to the molecular and neuronal mechanisms at work in the brain of a person living with a disorder of consciousness," explains Guillaume Dumas, who is also a professor at the Université de Montréal. "The model allows us to reproduce some of these phenomena and explore scenarios that would otherwise be impossible to study."
Better Understanding the Mechanisms of Recovery
The lesion simulations revealed a surprising finding: under certain patterns of brain injury, activating inhibitory neurons enabled the virtual brain to regain a much higher level of functioning than expected. Despite significant damage, the simulated neural circuits were able to reorganize themselves more efficiently.
These findings suggest that, in some cases, promoting a calming effect on brain activity may be more beneficial than attempting to stimulate it. In other words, helping the brain slow down could sometimes contribute more to recovery than trying to "wake it up."
"I was surprised by the extent of the recovery observed in the model," says Francesco Achilli, first author of the study. "As a psychiatrist, what interests me most is what these findings could mean for patients. Even partial recovery can represent a significant gain for a person living with a disorder of consciousness."
The findings suggest that inhibitory neurons play a more important role than previously believed in maintaining and restoring certain states of consciousness.
A Promising Tool for Research and Clinical Practice
Beyond its scientific findings, this study demonstrates the potential of computational modeling to advance research on disorders of consciousness. By enabling researchers to test different hypotheses virtually and explore the effects of various interventions on simulated brains, this approach could help explain why some individuals recover more successfully than others.
Although further validation is needed, these findings represent an important step toward a deeper understanding of the mechanisms underlying consciousness. In the long term, such models could contribute to the development of more personalized therapeutic approaches while helping bridge the gap between fundamental research and clinical practice for the benefit of people living with severe brain injuries.