Authors :
Alexander Andrews, Children's National; Dana Harrar, Children's National Health System; James Bost, Children's National Health System; Tesfaye G. Zelleke, Children's National Health System
Rationale:
Increased intracranial pressure (ICP) is common in the pediatric ICU after acquired brain injury. The gold standard for intracranial pressure measurement relies on invasive techniques. To date, there are no validated non-invasive techniques for intracranial pressure monitoring. A continuous, non-invasive measure of intracranial pressure based on EEG features would be of great value in the management of critically-ill children with acute brain injury.Quantitative EEG (qEEG) analysis can generate numerical values based on specific features of the EEG that can be compared between brain regions over time. To determine whether absolute values or changes in qEEG parameters might predict increases in intracranial pressure, we retrospectively applied qEEG to a 7 month-old with increased intracranial pressure due to a right middle cerebral artery stroke requiring craniectomy and implantation of an ICP monitor. We hypothesized that periods of increased ICP would be marked by a relative increase in the prevalence of delta (1-4 Hz) activity as determined by qEEG analysis.
Methods: Sustained (>2-hour) periods of normal ICP (<20mmHg) and high ICP (>20mmHg) were identified. Within these periods, 2-minute artifact-free segments of EEG spaced 15 minutes apart were visually selected for analysis. N-size is the number of ICP measurements. The dominant frequency (Hz) and average amplitude (mV) of the source EEG were evaluated within each 2-minute segment by an encephalographer (TZ) and neurology fellow (AA). EEG data was then processed using Persyst 13 (Persyst Development Corporation, Prescott, AZ) and subject to Fast Fourier Transformation.The power of the delta, theta, alpha, and beta frequency bands and the alpha (8-13Hz):delta (1-4Hz) power ratio were calculated for segments of normal ICP and segments of high ICP. Values were averaged over the whole head to account for hemispheric differences that might arise as a result of the unilateral stroke. Means were compared using the student's t-test. Results: EEG segments during times of normal ICP (<20mmHg) and times of increased ICP (>20mmHg) were differentiated by the alpha(8-13Hz):delta(1-4Hz) power ratio (mean) and 95% confidence interval 0.417 (0.381-0.454), 0.264 (0.240-0.289); (p-value <0.001). As anticipated, a low power ratio (relatively high prevalence of delta activity and/or low prevalence of alpha activity) correlated with increased ICP. Conclusions: A lower alpha:delta power ratio on EEG during times of increased intracranial pressure suggests that quantitative EEG may serve as an ancillary tool to detect increased intracranial pressure in the ICU, if these results are replicated in a larger study. This may, in turn, help to direct neuro-protective strategies in cases where invasive intracranial pressure monitoring is not feasible. Funding: No funding