Power Spectrum and Cross Power Spectral Density Based EEG Correlates of Intensive Care Delirium

Area of Science:

• Neuroscience
• Critical Care Medicine
• Medical Technology

Background:

• Delirium is a common and serious complication in intensive care unit (ICU) patients.
• Early detection and monitoring of delirium are crucial for improving patient outcomes.
• Electroencephalogram (EEG) monitoring offers a potential tool for objective delirium assessment.

Purpose of the Study:

• To identify specific electroencephalogram (EEG) correlates of delirium in ICU patients.
• To investigate the relationship between EEG signal characteristics and clinical delirium assessments.
• To explore the utility of EEG for monitoring delirium status in critical care settings.

Main Methods:

• Analysis of full montage EEG recordings from 15 ICU patients (23 recordings).
• EEG data analyzed in 10-second segments with 5-second overlap from 7 to 30 minutes.
• Relative power in conventional EEG frequency bands and absolute cross power spectral density were calculated at 20 electrode locations.

Main Results:

• Delirium was associated with lower relative power of EEG frequencies above 8 Hz, particularly in central and parietal regions.
• Significantly lower absolute cross power spectral density in the 20-30 Hz range was observed between electrode locations during delirium.
• These EEG changes correlated with Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) assessments.

Conclusions:

• Specific EEG patterns, including reduced high-frequency power and altered inter-electrode connectivity, are indicative of delirium in ICU patients.
• EEG analysis may serve as a valuable, objective tool for detecting and monitoring delirium in the ICU.
• Further research can refine EEG-based delirium detection algorithms for clinical application.