Global signal regression (GSR), a common preprocessing step in fMRI analysis, affects brain activity patterns differently depending on the anesthetic agent used. Using fMRI data from patients under general anesthesia, the work shows that GSR alters specific network connections under propofol but broadly reduces connectivity differences under sevoflurane. Network topology analyses reveal that GSR minimally affects propofol-induced changes in graph theoretical measures but significantly diminishes sevoflurane-related network alterations. These findings indicate that GSR's impact on functional brain organization is anesthetic-specific, with sevoflurane-induced changes being particularly sensitive to global signal removal. The results suggest that GSR should be applied cautiously when comparing different anesthetic agents.
Altered consciousness—from anesthesia and sleep to disorders of consciousness—involves distinct changes in thalamic nuclei. Analyzing fMRI data across these states, the authors found that propofol anesthesia disrupted pulvinar-cortical connections, sleep transitions affected specific nuclei (VLp, medial geniculate, centromedian), and disorders of consciousness showed widespread disconnections. Five key nuclei showed state-specific alterations, with higher-order nuclei (pulvinar, centromedian, mediodorsal) more consistently involved. Decreased local brain signal complexity occurred in 4–6 nuclei during anesthesia and 4–5 in patients with disorders of consciousness. The coupling between local fluctuation and connectivity varied systematically with consciousness state, suggesting potential therapeutic targets.