Spectral Cerebral Blood Volume Accounting for Noninvasive Estimation of Changes in Cerebral Perfusion Pressure in Patients with Traumatic Brain Injury.

Abstract

We present the application of a new method for non-invasive cerebral perfusion pressure estimation (spectral nCPP or nCPPs) accounting for changes in transcranial Doppler-derived pulsatile cerebral blood volume. Primarily, we analysed cases in which CPP was changing (delta [∆],magnitude of changes]): (1) rise during vasopressor-induced augmentation of ABP (N=16); and (2) spontaneous changes in intracranial pressure (ICP) during plateau waves (N=14). Secondarily, we assessed nCPPs in a larger cohort in which CPP presented a wider range of values. The average correlation in the time domain between CPP and nCPPs for patients undergoing an induced rise in arterial blood pressure (ABP) was 0.95±0.07. For the greater traumatic brain injury (TBI) cohort, this correlation was 0.63±0.37. ∆ correlations between mean values of CPP and nCPPs were 0.73 (p=0.002) and 0.78 (p<0.001) respectively for induced rise in ABP and ICP plateau wave cohorts. The area under the curve (AUC) for ∆CPP was of 0.71 with a 95% confidence interval of 0.54-0.88. To detect low CPP, AUC was 0.817 with a 95% confidence interval of 0.79-0.85. nCPPs can reliably identify changes in direct CPP across time and the magnitude of these changes in absolute values. The ability to detect changes in CPP is reasonable but stronger for detecting low CPP, ≤70mmHg.