What does this research mean for the field?

The pulse shape index (PSI) is a valid automated alternative to the P2/P1 ratio for assessing intracranial compliance, though optimal monitoring requires combining morphological metrics with amplitude and reserve indices. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The aim is to validate the pulse shape index (PSI) against the P2/P1 ratio for assessing intracranial compliance.

August 5, 2026Neurocritical Care0 citationsOpen Access

Waveform-Derived Metrics of Intracranial Compliance: Validation and Pulsatile Dynamics

IOIhsane OlakoredeUniversity of Cambridge SBStefan Yu BögliUniversity of Zurich MCMarek CzosnykaWarsaw University of Technology

Key Points

The aim is to validate the pulse shape index (PSI) against the P2/P1 ratio for assessing intracranial compliance.
Analyzed 4388 segments from 96 high-resolution ICP and TCD recordings.
Calculated PSI, P2/P1, ICP pulse amplitude, and compliance indices for each segment.
Used correlation, mixed-effects models, ROC, and precision-recall analyses for evaluation.
PSI showed a strong association with P2/P1 (r ≈ 0.74–0.76) and high discrimination for impaired compliance (AUC ≈ 0.93).
RAP and intracranial compliance indices had weak associations with P2/P1 and showed substantial variability.
Low ICP pulse amplitude (< 1.5 mm Hg) correlated with impaired compliance and lower RAP.

Abstract

Abstract Objective Visual interpretation of intracranial pressure (ICP) waveforms is widely used to assess intracranial compliance but lacks rigorous validation. We evaluated the pulse shape index (PSI) against the tidal wave (P2)/percussion wave (P1) ratio and examined the contributions of pulse amplitude, compensatory reserve pressure–volume compensatory reserve index (RAP) and transcranial Doppler (TCD)-derived indices. Methods A total of 4388 non-overlapping 5-min segments from 96 high-resolution recordings of simultaneous ICP and TCD blood velocity were analysed. PSI and P2/P1 were derived from normalised ICP pulses. ICP pulse amplitude (AMP), RAP and TCD-derived arterial and intracranial compliance indices arterial compliance (C A ), intracranial compliance (C I ) were calculated per segment. Associations were assessed at segment, recording, and patient levels using correlation and mixed-effects models. Discriminatory performance for impaired (P2/P1 > 1.1) and preserved (P2/P1 < 0.9) compliance was evaluated using receiver operating characteristic (ROC) and precision–recall analyses. Concordance between PSI and P2/P1 classifications was assessed using Chi-squared threshold mapping. Results PSI demonstrated a strong association with P2/P1 ( r ≈ 0.74–0.76 across analysis levels) and high discrimination for impaired compliance area under the curve (AUC) ≈ 0.93. RAP and C I showed weak and inconsistent population-level associations with P2/P1, with substantial inter-individual variability. A distinct phenotype of low AMP (< 1.5 mm Hg) despite impaired waveform-derived compliance was associated with lower RAP. C A and C I provided limited incremental explanatory value, whereas inclusion of AMP improved model fit and classification performance beyond morphology-based metrics alone. Conclusions PSI closely reflects P2/P1 and supports automated waveform-based compliance assessment. However, morphology alone is insufficient; amplitude and reserve indices provide complementary information, supporting a multiparametric approach to intracranial monitoring.

KI fragen

Bookmark

View Full Paper