Functional hub disruption emphasizes consciousness recovery in severe traumatic brain injury

Severe traumatic brain injury can lead to transient or even chronic disorder of consciousness. To increase diagnosis and prognosis accuracy of disorder of consciousness, functional neuroimaging is recommended one month post injury. Here, we investigated brain networks remodelling on longitudinal data between one- and three-months post severe traumatic brain injury related to change of consciousness. Thirty-four severe traumatic brain injured patients were included in a cross sectional and longitudinal clinical study and their MRI data were compared to those of twenty healthy subjects. Long duration resting-state functional MRI were acquired in minimally conscious and conscious patients at two time points after their brain injury. The first time corresponds to the exit from Intensive Care Unit and the second one to the discharge from post intensive care rehabilitation ward. Brain networks data were extracted using graph analysis and metrics at each node quantifying local (Clustering) and global (Degree) connectivity characteristics. Comparison with brain networks of healthy subjects revealed patterns of hyper- and hypo- connectivity that characterize brain networks reorganization through the hub disruption index, a value quantifying the functional disruption in each individual severe traumatic brain injury graph. At discharge from intensive care unit, twenty-four patients’ graphs (nine minimally conscious and fifteen conscious) were fully analysed and demonstrated significant network disruption. Clustering and Degree nodal metrics, respectively related to segregation and integration properties of the network, were relevant to distinguish minimally conscious and conscious groups. At discharge from post intensive care rehabilitation unit, fifteen patients’ graphs (two minimally conscious, thirteen conscious) were fully analysed. The conscious-group still presented a significant difference with healthy subjects. Using mixed effects models, we showed that consciousness state, rather than time, explained the Hub disruption index differences between minimally conscious and conscious-groups. While severe traumatic brain injured patients recovered full consciousness, regional functional connectivity evolved towards a healthy pattern. More specifically the restoration of a healthy brain functional segregation could be necessary for consciousness recovery after severe traumatic brain injury. For the first time, extracting the Hub disruption index directly from each patient’s graph, we were able to track the clinical alteration and subsequent recovery of consciousness during the first three months following a severe traumatic brain injury.