A practical measure of integrated information reveals alpha-band activity and the posterior cortex as neural correlates of arousal.
NeuroImage July 18, 2025 Xin Wen, Yu Chang, Sijie Li et al. 1 citation
A new measure called Φcopula, which uses a Gaussian copula approach to estimate integrated information, outperforms common estimators by maintaining the lowest bias and mean squared error even in non-Gaussian high-dimensional systems. Applied to electroencephalographic data across awake, propofol-induced unresponsive, and NREM sleep states, alpha-band Φcopula significantly decreased during both anesthesia and sleep. Φcopula-based classifiers distinguished arousal states more accurately than functional connectivity and network efficiency measures. The dorsal attention network and default mode network contributed most to Φcopula, with the cingulate and posterior cortices showing the greatest contributions. The posterior cortex, especially the posterior cingulate cortex, appears critical for arousal-related information integration and consciousness.