A systematic investigation of the association between network dynamics in the human brain and the state of consciousness

Brain dynamics measured with resting-state fMRI are linked to the state of consciousness, not just to behavioral responsiveness. A longitudinal study of patients recovering from coma assessed patients twice, six months apart. One group regained consciousness and improved responsiveness; the other was already conscious and only improved responsiveness. At the first assessment, conscious patients showed greater variance in network metrics than unconscious patients, a difference that disappeared once all patients had recovered consciousness. Patients who regained consciousness showed a significant increase in dynamics over time, while those who only improved responsiveness did not. This suggests changes in brain dynamics relate specifically to consciousness.

Study at a glance

Abstract

Abstract An increasing amount of studies suggest that brain dynamics measured with resting-state functional magnetic resonance imaging (fMRI) are related to the state of consciousness. However, the challenge of investigating neuronal correlates of consciousness is the confounding interference between (recovery of) consciousness and behavioral responsiveness. To address this issue, and validate the interpretation of prior work linking brain dynamics and consciousness, we performed a longitudinal fMRI study in patients recovering from coma. Patients were assessed twice, 6 months apart, and assigned to one of two groups. One group included patients who were unconscious at the first assessment but regained consciousness and improved behavioral responsiveness by the second assessment. The other group included patients who were already conscious and improved only behavioral responsiveness. While the two groups were matched in terms of the average increase in behavioral responsiveness, only one group experienced a categorical change in their state of consciousness allowing us to partially dissociate consciousness and behavioral responsiveness. We find the variance in network metrics to be systematically different across states of consciousness, both within and across groups. Specifically, at the first assessment, conscious patients exhibited significantly greater variance in network metrics than unconscious patients, a difference that disappeared once all patients had recovered consciousness. Furthermore, we find a significant increase in dynamics for patients who regained consciousness over time, but not for patients who only improved responsiveness. These findings suggest that changes in brain dynamics are indeed linked to the state of consciousness and not just to a general level of behavioral responsiveness.