Systemic cardiovascular inputs in models estimating intracranial pressure magnitude and waveform

Background: Monitoring Intracranial Pressure (ICP) is key for appropriate clinical treatment of patients with conditions potentially causing raised ICP. The adequacy of using Heart Rate (HR), aortic Blood Pressure (aBP) and carotid Blood Flow (cBF) to estimate ICP magnitude (pulse and mean) and waveform is investigated as an alternative means to invasive ICP measurement. Methods: ICP (sequentially raised from resting ICP to 30–40 mmHg with infusions of artificial intracranial fluid), aBP (lowered with sodium nitroprusside and raised with phenylephrine, 30 μg/kg/min, across a physiological range), HR (paced at 400 and 500 bpm), and cBF were measured in 11 anaesthetised Sprague Dawley rats. Potential cardiovascular predictors of ICP magnitude were assessed by stepwise mixed-model regression. Two transfer function models were constructed to estimate the ICP waveform from aBP or cBF waveforms. Results: Systolic, mean and diastolic aBP as well as peak and minimum cBF had significant predictive value for mean ICP (p < 0.001, R2 = 0.25). HR (p < 0.05), systolic and mean aBP (p < 0.001), peak (p < 0.001), mean (p < 0.05) and minimum (p < 0.01) cBF had significant value for pulse ICP (R2 = 0.35). The transfer function models showed potential to reproduce the ICP waveform (Root Mean Square Error (RMSE)≤4 mmHg), being more accurate for mean aBP above 100 mmHg and mean ICP below 20 mmHg (RMSE ≤ 0.5 mmHg). Conclusions: The models developed from the comprehensive rat experiment demonstrated that systemic cardiovascular measures have predictive value in estimating the ICP magnitude and waveform, but other inputs may be necessary to improve accuracy in estimating ICP across the full physiological range.