Conference Agenda

This study investigated auditory processing during propofol sedation in healthy participants by recording brain activity with fMRI during an oddball paradigm - where standard, repetitive sounds were occasionally interrupted by deviant ones. At the end of each of the two auditory stimulation sessions, participants were awakened to report their state of consciousness and sound perception during the preceding unresponsive period. Reports of dreaming with sound perception were classified as indicating connected consciousness (CC), while reports of dreaming without sound perception as disconnected consciousness (DC).

The final dataset included 34 sedation sessions (18 CC, 16 DC) from 25 participants. Propofol concentrations did not differ significantly between CC and DC across sessions (all p-values > 0.2). Cluster-corrected fMRI analyses (FWE-corrected at p<0.05, cluster-defining threshold p<0.001) revealed greater activation in DC than CC in response to all sounds and standard tones, notably in the superior/middle temporal and precentral gyri, possibly reflecting feedforward error propagation. Conversely, CC participants showed widespread decreases with focal increases in the precuneus and, at a more lenient threshold (z ±1.96, p<0.05), in prefrontal, hippocampal, and occipital regions. For deviant sounds, this pattern reversed: CC participants showed widespread activations compared to DC, specifically in the right fusiform, right lingual, and left precentral gyri (z ±3.29, p<0.001), embedded in a broader fronto-parietal and temporal network at a more lenient threshold (z ±1.96, p<0.05). Conversely, DC participants showed minimal differences between deviant and standard sound processing, suggesting deviant perception may be exclusive to CC.

Background: Ketamine, in doses suitable to induce anaesthesia in humans, gives rise to a unique state of unresponsiveness accompanied by vivid experiences and sensations, making it possible to disentangle the correlated but distinct concepts of conscious awareness and behavioural responsiveness. This is a distinction that is often overlooked in the study of consciousness.

Methods: The mathematical framework of connectome harmonic decomposition (CHD) was used to view functional MRI signals during ketamine-induced unresponsiveness in terms of distributed patterns across spatial scales. Afterwards, the connectome harmonic signature of this particular state was mapped onto signatures of various other states of consciousness.

Results: An increased prevalence of fine-grained connectome harmonics was found in functional MRI signals obtained during ketamine-induced unresponsiveness, indicating higher granularity. After statistical assessment, ketamine sedation’s harmonic signature showed alignment with signatures of LSD-induced or ketamine-induced psychedelic states and misalignment with those seen in unconscious individuals, whether due to propofol sedation or brain injury.

Conclusions: The CHD framework, which only requires resting-state fMRI data and can be applied retrospectively, has the ability – at least on a group level – to reliably track alterations in conscious awareness in the absence of behavioural responsiveness. This discovery was made possible by ketamine’s unique property of decoupling these two facets, and is of crucial importance for consciousness and anaesthesia research.

Introduction:

Understanding the neural correlates of consciousness is a fundamental scientific question, relevant to understanding anaesthesia, coma, delirium, and psychosis. While the brainstem plays a key role in this context, the resolution of functional MRI (fMRI) at 3 Tesla (3T) has largely confined studies to group-level analyses, which cannot address subject-specific changes. We present data showing how the increased resolution of 7T allows for analysis of single-subject changes during administration of anaesthesia.

Methods:

We have undertaken a pilot study using 7T fMRI to interrogate brainstem connectivity in two subjects during and after an infusion of the anaesthetic agent, propofol, to the point of loss of responsiveness, as assessed by responses to an auditory discrimination task. Utilising the high spatiotemporal resolution and signal-to-noise-ratio of 7T fMRI we assessed for single subject changes to brainstem connectivity throughout anaesthetic induced transitions of consciousness.

Results:

Higher plasma propofol concentrations were associated with slower reaction times (subject 1: F-value=6.11, p<0.001, subject 2: F-value=10.2, p<0.001) followed by loss of responsiveness. Higher plasma propofol concentrations were also associated with an increase in intra-brainstem connectivity (subject 1: F-value=12.7, p<0.001; subject 2: F-value=16.5, p<0.001), with changes reverting to baseline on recovery from anaesthesia. Post-hoc analysis showed increases in intra-brainstem connectivity were specific to periods of decreased behavioural responsiveness (p<0.001).

Conclusion:

Our pilot study suggests anaesthetic induced loss of responsiveness is associated with a marked but reversible increase in intra-brainstem connectivity, with brainstem arousal nuclei seemingly losing their functional independence. A larger study will be undertaken to verify these findings.

The neurobiology of consciousness remains one of neuroscience's greatest challenges. Pharmacological interventions that alter arousal provide valuable tools for investigating how the brain generates conscious experiences—if their effects are specific to consciousness. However, we have recently shown that the sedative Propofol (Fontan et al., 2021) influences also unconscious neural processing, challenging the assumption that reduced arousal selectively affects consciousness. Here, we explore whether this non-specificity extends to noradrenergic modulation, aiming to further clarify the relationship between the "level" and "content" of consciousness.

In Study 1, we assessed the effects of Dexmedetomidine (Dexdor), a selective α2A adrenergic receptor agonist, on BOLD signals during conscious and unconscious visuospatial processing. Study 2 used sleep deprivation as a natural model of altered arousal, which qualitatively parallels Dexdor in its effects on arousal. Both Dexdor sedation and sleep deprivation selectively influenced conscious processing, as indicated by changes in fMRI BOLD signals, suggesting a specific role for noradrenaline in consciousness.

Interestingly, the altered spatial biases during conscious trials differed between the two conditions. Moderate Dexdor sedation reduced leftward bias, while sleep deprivation increased it. The latter likely reflects compensatory mechanisms during sleep deprivation, where increased heart rates indicated participants efforts to stay awake, supported by a positive correlation between heart rate and leftward bias.

Our results provide the first pharmacological neuroimaging evidence of noradrenaline’s involvement in conscious visuospatial processing, highlighting the intricate interaction between level and content in shaping consciousness.

Introduction: Zolpidem is a GABAergic non benzodiazepine hypnotic drug used as a treatment of short-term insomnia. Case reports in the literature have reported paradoxical awakening effect on patients with disorder of consciousness (DoC), with an inconsistent response. We hypothesized that the response variability could be explained by lesional topography.

Methods : We conducted a longitudinal retrospective study in 91 DoC patients : 61 acute patients in intensive care unit and 30 chronic patients in rehabilitation unit. The effect of zolpidem was assessed by CRS-R before and after 10mg of zolpidem. Lesion localization was performed on morphological MRI by 2 blinded raters with dissensus resolution by a 3rd rater.

Results: Only 6 chronic DoC patients exhibited a paradoxical effect; these patients demonstrated bilateral preservation of striato-pallidal complex.

Conclusion: Our results suggest that the paradoxical effect of zolpidem is mediated by plasticity mechanisms occurring in the preserved brain network including the striato-pallidal complex . We hypothesize that this paradoxical effect may involve plasticity in medium aspiny neurons, which are GABAergic neurons targeted by zolpidem and play a role in the striatal microcircuitry. This study is unique due to the size of the cohort and the inclusion of both acute and chronic patients. Further investigations are needed to explore these mechanisms.

General anaesthesia is widely assumed to produce unconsciousness, yet a growing body of evidence indicates that some patients can remain connected to their environment whilst being behaviourally unresponsive. This study employed the isolated forearm technique (IFT) and functional MRI (fMRI) to investigate the neural correlates of connected consciousness under propofol anaesthesia. Twenty-six healthy volunteers received moderate doses of propofol and remifentanil. After achieving behavioural unresponsiveness, an individually adjusted noxious electrical stimulus was applied to the sural nerve, followed by delivering verbal commands to assess behavioural responsiveness. Resting state fMRI sequences were collected during normal wakefulness, unresponsiveness-induced anaesthesia, after noxious stimulation, and recovery (i.e., after anaesthesia stopped). Forty percent (10/25) of participants subsequently regained behavioural responsiveness following the noxious stimulus, thus were deemed responders. Unresponsiveness under moderate propofol anaesthesia was not associated with a widespread cortical suppression. Instead, there were increases in the functional connectivity (FC) of primary sensory regions and decreases in the FC of several secondary associated regions supporting internetwork connectivity. Notably, the transition to behavioural responsiveness was pre-emptively enabled by enhanced thalamic network integrity and maintained through increased FC between the default mode network and thalamus and the auditory network and hippocampus, whilst non-responders had higher connectivity between the executive control network and thalamus. These findings underscore that behavioural unresponsiveness under anaesthesia does not necessarily equate to unconsciousness. This work begins to disentangle the neural correlates of unresponsiveness from those of unconsciousness and clinically underscores the importance of identifying intraoperative connected consciousness to minimise potential suffering.

General anaesthesia induces a drastic breakdown of information processing in the brain, suppressing both sensory and motor interactions between organism and environment, and ultimately consciousness.

Here, we investigate the neural mechanisms of anaesthesia across humans, non-human primates, and rodents, integrating functional neuroimaging with bi-directional causal manipulations through pharmacology and intracranial stimulation. Using a rigorous quantification of integrated information based on the information-theoretic notion of synergy, we show that diverse anaesthetics induce a breakdown of integrated information in functional MRI signals, across human, macaque, and mouse. Integrated information is then consistently restored upon re-awakening induced by thalamic deep brain stimulation in the macaque. Across all three species, the regional disruption of integrated information is spatially correlated with species-specific expression of PVALB/Pvalb gene, a cell-type marker for inhibitory interneurons. To provide mechanistic insights, we reproduce the effects of anaesthesia in biophysical computational models integrating species-specific connectomics and transcriptomics for human, macaque, and mice. Our models reveal that the regional expression of parvalbumin gene is especially suitable for controlling the integration of information via regionally heterogeneous inhibition, whereas the structural connectivity of the central thalamic nucleus in the macaque brain makes it especially suitable as a stimulation target for restoring integration of information – replicating our empirical results. Overall, our multi-modal, multi-species investigation reveals how network structure and neuromodulation jointly orchestrate information integration and consciousness in the mammalian brain. These findings have far-reaching implications for clinical practice, given the relevance of deep brain stimulation as potential treatment for patients with disorders of consciousness.