Introduction
To both evaluate cerebrovascular autoregulation and provide on-going insight into cerebrovascular pathophysiology of patients with brain injury we have been exploring the development of a numerical identification modeling technique which uses arterial blood pressure (ABP) and intracranial pressure (ICP) recordings. This numerical technique requires an assumed mathematical structure for the physical process of cerebrovascular pressure transmission. We have based our assumption on a previously proposed analog circuit model of ICP dynamics. Just as the modal frequencies, critical vibration modes, of a tuning fork reflect its structural properties; the modal frequencies of cerebrovascular pressure transmission reflect the structural properties of the cerebrovascular bed. To conveniently quantify cerebrovascular pressure transmission, we chose to evaluate the highest modal frequency (HMF) which represents the highest critical vibration frequency of transmission. The purpose of this study was to test the validity of our mathematical assumption required for numerical identification modeling by comparing the simulations of the analog model with those obtained by identification modeling technique generated from the same clinical pressure recordings.
Methods
Pressure recordings were obtained from 6 patients with traumatic brain injury, a group (n=3) with mean cerebral perfusion pressure (CPP) >50 mmHg and a group (n=3) with mean CPP <30 mmHg. Values of HMF were computed by two methods: 1) derivation by the numerical identification autoregressive moving average technique; and 2) derivation from the analog circuit model of ICP dynamics. For each simulation we used a clinical ABP recording as the applied pressure source and manipulated the parameters of the analog model to obtain a match of the: 1) simulated and numerically derived values of HMF within 0.3 Hz; and 2) clinical and simulated ICP recordings.
Results
All simulations of ICP generated by the analog circuit model visually match the corresponding clinical ICP recording. The group of patients with CPP >50 mmHg were assumed to have intact autoregulation with a mean cerebral blood flow (CBF) at approximately 750 ml/min. The other patients were assumed to have impaired autoregulation with a significantly lower mean CBF. Simulated values of intracranial compliance were found to increase with cerebral hypoperfusion. Simulated mean values of CPP, HMF, and intracranial compliance (IC), and CBF are given in Table 1.
Summary of Means (± S.D.) of Simulated HMF, Arteriol. Resist., IC, and CBF
Conclusions
The findings of this study are: 1) simulations of ICP produced by the analog circuit model match clinical ICP recordings for equivalent values of simulated and numerically derived HMF; and 2) the simulated value of IC increases during impaired regulation with cerebral hypoperfusion. These findings support the validity of the mathematical assumption required for the proposed numerical identification modeling technique of cerebrovascular pressure transmission.