Information sharing in the brain indexes consciousness in noncommunicative patients

Abstract

Neuronal theories of conscious access tentatively relate conscious perception to the integration and global broadcasting of information across distant cortical and thalamic areas. Experiments contrasting visible and invisible stimuli support this view and suggest that global neuronal communication may be detectable using scalp electroencephalography (EEG). However, whether global information sharing across brain areas also provides a specific signature of conscious state in awake but noncommunicating patients remains an active topic of research. We designed a novel measure termed "weighted symbolic mutual information" (wSMI) and applied it to 181 high-density EEG recordings of awake patients recovering from coma and diagnosed in various states of consciousness. The results demonstrate that this measure of information sharing systematically increases with consciousness state, particularly across distant sites. This effect sharply distinguishes patients in vegetative state (VS), minimally conscious state (MCS), and conscious state (CS) and is observed regardless of etiology and delay since insult. The present findings support distributed theories of conscious processing and open up the possibility of an automatic detection of conscious states, which may be particularly important for the diagnosis of awake but noncommunicating patients.

Figure 1. Weighted Symbolic Mutual Information

(A) The transformation of continuous signals (X) into sequences (X̂) of discrete symbols (A, B, … F) enables an easy and robust estimation of the mutual information shared between two signals. The τ parameter refers to the temporal separation of the elements that constitute a symbol, composed of three elements. (B) By computing the joint probability of each pair of symbols, we can estimate the symbolic mutual information (SMI) shared across two signals. (C) To compute weighted symbolic mutual information (wSMI), the SMI is weighted to disregard conjunctions of identical or opposite-sign symbols, which could potentially arise from common-source artifacts.

Figure 2. wSMI Indexes Consciousness Independently of Etiology and Delay since Insult

The median wSMI across current sources is depicted for each state of consciousness (A). Error bars represent SEM. Significant pairwise comparisons are denoted with asterisks. Analyses were reproduced for each etiology (B) and delay since insult (C and D). The results showed that median wSMI is mainly affected by the state of consciousness and does not vary significantly across etiology or delays.

Figure 3. wSMI Increases with Consciousness, Primarily over Centroposterior Regions

(A) The median wSMI that each EEG channel shares with all other channels is depicted for each state of consciousness. (B) 120 pairs formed by 16 clusters of EEG channels are depicted as 3D arcs whose height is proportional to the Euclidian distance separating the two clusters. Line color and thickness are proportional to the mean wSMI shared by the corresponding cluster pair.

Figure 4. wSMI as Function of Interchannel Distance

(A) wSMI is plotted as a function of the Euclidian distance separating each pair of EEG channels. While wSMI is relatively stable between 8 and 23 cm, it drops toward zero as interchannel distances diminish, which thus confirms its robustness to common source artifacts. (B) Histogram plotting the density of channel pairs as a function of interchannel distance. (C) VS patients presented lower wSMI than MCS and CS patients, particularly over medium and long interchannel distances (>10 cm).