Dissociations between spontaneous EEG features and the Perturbational
Complexity Index in the minimally conscious state
Abstract
The analysis of spontaneous EEG is a cornerstone in the assessment of
patients with disorders of consciousness (DoC). Alterations in specific
frequency bands have been reported, including a predominance of delta
power in vegetative state/unresponsive wakefulness syndrome (VS/UWS)
patients in contrast with a predominance of alpha activity in minimally
conscious state patients (MCS). Although preserved EEG patterns are
highly suggestive of consciousness even in unresponsive patients,
moderately or severely abnormal patterns are difficult to interpret.
Indeed, growing evidence shows that consciousness can be present despite
either large delta or reduced alpha activity in spontaneous EEG.
Quantifying the complexity of EEG responses to direct cortical
perturbations (Perturbational Complexity Index; PCI) may complement the
observational approach and provide a reliable assessment of
consciousness even when spontaneous EEG features are inconclusive. To
systematically test this hypothesis, we compared PCI to EEG spectral
measures in the same population of MCS patients (n=40) hospitalized in
rehabilitation facilities. We found a remarkable EEG background
variability across MCS patients as compared to healthy controls and a
non-negligible number of patients with predominant delta and highly
reduced alpha power in spontaneous EEG. Conversely, PCI values
invariably suggested a capacity for consciousness in all MCS patients,
consistent with the presence of clearly discernible, albeit fleeting,
behavioral signs of awareness. These results confirm that spontaneous
EEG rhythms may dissociate from the actual capacity for consciousness
and suggest that a perturbational approach can effectively compensate
for this pitfall with practical implications for the individual
patient’s stratification and tailored rehabilitation.