Noninvasive Intracranial Pressure Determination in Patients with Subarachnoid Hemorrhage

Intracranial pressure (ICP) should ideally be measured in many conditions affecting the brain. The invasiveness and associated risks of the measurement modalities in current clinical practice restrict ICP monitoring to a small subset of patients whose diagnosis and treatment could benefit from ICP measurement. To expand validation of a previously proposed model-based approach to continuous, noninvasive, calibration-free, and patient-specific estimation of ICP to patients with subarachnoid hemorrhage (SAH), we made waveform recordings of cerebral blood flow velocity in several major cerebral arteries during routine, clinically indicated transcranial Doppler examinations for vasospasm, along with time-locked waveform recordings of radial artery blood pressure (APB), and ICP was measured via an intraventricular drain catheter. We also recorded the locations to which ICP and ABP were calibrated, to account for a possible hydrostatic pressure difference between measured ABP and the ABP value at a major cerebral vessel. We analyzed 21 data records from five patients and were able to identify 28 data windows from the middle cerebral artery that were of sufficient data quality for the ICP estimation approach. Across these windows, we obtained a mean estimation error of -0.7 mmHg and a standard deviation of the error of 4.0 mmHg. Our estimates show a low bias and reduced variability compared with those we have reported before.