Multidimensional assessment of heartbeat-evoked responses in disorders of consciousness.

Because consciousness does not necessarily translate into overt behaviour, detecting residual consciousness in noncommunicating patients remains a challenge. Bedside diagnostic methods based on EEG are promising and cost-effective alternatives to detect residual consciousness. Recent evidence showed that the cortical activations triggered by each heartbeat, namely, heartbeat-evoked responses (HERs), can detect through machine learning the presence of minimal consciousness and distinguish between overt and covert minimal consciousness. In this study, we explore different markers to characterize HERs to investigate whether different dimensions of the neural responses to heartbeats provide complementary information that is not typically found under standard event-related potential analyses. We evaluated HERs and EEG average non-locked to heartbeats in six types of participants: healthy state, locked-in syndrome, minimally conscious state, vegetative state/unresponsive wakefulness syndrome, comatose and brain-dead patients. We computed a series of markers from HERs that can generally separate the unconscious from the conscious. Our findings indicate that HER variance and HER frontal segregation tend to be higher in the presence of consciousness. These indices, when combined with heart rate variability, have the potential to enhance the differentiation between different levels of awareness. We propose that a multidimensional evaluation of brain-heart interactions could be included in a battery of tests to characterize disorders of consciousness. Our results may motivate further exploration of markers in brain-heart communication for the detection of consciousness at the bedside. The development of diagnostic methods based on brain-heart interactions may be translated into more feasible methods for clinical practice.