Permutation Entropy Analysis to Intracranial Hypertension from a Porcine Model.

Area of Science:

• Neurology
• Biomedical Engineering
• Data Science

Background:

• Intracranial pressure (ICP) monitoring is crucial in intensive care but often underutilizes available time-series data.
• Intracranial compliance is a key metric for patient management, yet its direct measurement from ICP is challenging.

Purpose of the Study:

• To introduce permutation entropy (PE) as a method for extracting non-obvious information from ICP time series.
• To investigate the relationship between PE, number of missing patterns (NMP), and intracranial compliance.
• To establish potential indicators for altered neurophysiology using PE and NMP.

Main Methods:

• Analysis of ICP data from a pig experiment using sliding windows.
• Estimation of permutation entropy (PE), probability distributions, and number of missing patterns (NMP).
• Correlation analysis between PE, NMP, and ICP dynamics.

Main Results:

• Permutation entropy (PE) exhibits an inverse relationship with ICP.
• The number of missing patterns (NMP) serves as a surrogate for intracranial compliance.
• Specific PE and NMP values indicate lesion-free states versus terminal lesion phases, suggesting potential for early warnings.

Conclusions:

• Permutation entropy (PE) and NMP provide valuable, non-obvious insights into ICP dynamics and intracranial compliance.
• Deviations from established PE and NMP thresholds may signal altered neurophysiology.
• These methods show promise for real-time patient monitoring and as input for machine learning in critical care.