A pilot study for continuous non-invasive intracranial pressure screening

Intracranial hypertension (ICH) is a common and critical condition in neurocritical care, often requiring immediate intervention. Current methods for continuous intracranial pressure (ICP) monitoring are invasive and costly, limiting their use in resource-limited settings. This study investigates the potential of the electroencephalography (EEG) as a non-invasive alternative for ICP monitoring. Using a generalized additive model (GAM) applied to porcine models, we analyzed the correlation between ICP and EEG features, including total EEG energy and the slope of the EEG power spectral density. Our model achieved moderate specificity (0.75) and sensitivity (0.73), accurately identifying 85% of ICH episodes with a 3-second window. These findings support the feasibility of using the EEG to detect ICH, particularly in low-resource environments. While further validation in human subjects is needed, our approach offers a promising, cost-effective method for non-invasive continuous ICP monitoring, enhancing ICH detection where traditional methods are impractical or inaccessible.