Connectivity and Neural Signatures of Consciousness in Unresponsive States

Individuals with severe brain injury often require extensive treatment in intensive care units (ICU) and hospitalization wards while uncertainty prevails about the recovery of consciousness and cognitive abilities. Especially in the acute phase after injury, treatment decisions have a tremendous impact on outcome, but rely on assessments of behavioral responsiveness which are known to be unreliable and subject to many confounders. Objective, quantifiable diagnostic and prognostic measures that can be deployed at the bedside during the ICU stay are lacking. Development of new metrics are hampered by our lack of a fundamental understanding of (i) thalamocortical network mechanisms underlying consciousness and (ii) brain-injury induced neural dynamics impacting both consciousness and outcome. Continuous EEG monitoring has been used to aid in this respect to (i) predict recovery of consciousness and outcome, and (ii) diagnose nonconvulsive seizures in unresponsive patients. Although promising, it lacks sensitivity, spatial resolution, and causal power. There is an urgent need for techniques allowing high-precision detection of pathological dynamics, patient stratification and prediction of a capacity for consciousness recovery in acute unresponsive patients with brain injury. Intracranial electrodes as part of multimodal monitoring in subjects with impaired consciousness and severe brain injury allow continuous bedside recordings of high spatiotemporal resolution in different network nodes and allows inducing brain perturbations, transcending correlational evidence of network analysis. This technique could increase the detection and treatment of nonconvulsive seizures contributing to brain injury and unresponsiveness and simultaneously allows to study networks supporting consciousness. This can lead to new diagnostic and prognostic biomarkers for recovery based on thalamocortical profiles of activity, reactivity (complexity) and connectivity, ultimately paving to way for the development of biomarker-driven treatments to support early recovery such as deep brain stimulation. Eventually this can contribute to clinical decision making: abstaining from aggressive treatment in patients with no potential for recovery, and more importantly, continuing treatment in subjects with a responsive brain but without clear behavioural correlate.