System-specific reductions of resting-state BOLD signal entropy in disorders of consciousness

Disorders of consciousness (DoC) are associated with large-scale abnormalities in brain function, but system-specific alterations in regional resting-state BOLD signal entropy remain poorly understood. Here, we examined regional entropy of resting-state fMRI signals in patients with minimally conscious state (MCS) and unresponsive wakefulness syndrome (UWS).

Resting-state fMRI was acquired in 23 MCS patients, 31 UWS patients, and 20 age-matched healthy controls. Regional entropy was estimated from voxel-wise BOLD signals within anatomically defined brain regions using a PCA-based framework, allowing quantification of covariance-based regional BOLD signal variability across the whole brain. Pairwise group comparisons were performed at both the ROI level and in grouped functionally relevant regional sets.

Relative to healthy controls, MCS patients showed reduced regional entropy mainly in sensory- and memory-related regions, whereas UWS patients showed widespread reductions across sensory, memory-related, and high-order cognitive regions. In grouped regional-set analyses, entropy in sensory and memory systems showed a graded decrease from healthy controls to MCS and from MCS to UWS. In contrast, entropy in high-order cognitive systems remained relatively preserved in MCS and was significantly reduced only in UWS, with the exception of the posterior cingulate cortex and precuneus.

These findings demonstrate system-specific reductions in resting-state BOLD signal entropy in DoC and reveal distinct patterns of altered regional BOLD signal variability across sensory, memory-related, and high-order cognitive systems in MCS and UWS. Regional BOLD signal entropy may provide a quantitative regional description of altered brain dynamics in DoC and complement existing work on large-scale communication and network dysfunction.