Original ContributionImproved Parameterization of the Transcranial Doppler Signal
Introduction
Transcranial Doppler (TCD) can theoretically contribute to the monitoring of critically ill patients in whom brain circulation is at stake (Aaslid et al. 1982; Bhatia and Gupta 2007). However, despite, the well defined relation between pulsatility index (PI) and intracranial pressure (ICP) (e.g., Bellner et al. 2004), TCD on its own has not achieved a prominent position on (neuro-) intensive care (ICU). TCD in combination with a continuous registration of arterial pressure (ABP) and intracranial pressure (ICP) is still under investigation (e.g., Czosnyka et al. 2003). In most ICUs, therapy remains restricted to optimization of systemic ABP and breathing parameters without detailed information about its effects on brain circulation.
To improve the capability of TCD for neurovascular monitoring, two important problems need to be addressed. First, performing a TCD investigation is operator dependent. This problem is addressed by self-searching TCD probes currently being introduced to the market. Second, the interpretation of the TCD signal depends on a great number of technical and physiologic factors contributing to the signal in various ways.
Regarding the technical factors, it should be emphasized that TCD measures flow velocity (FV) and not flow. Therefore, the measurements are expressed in meters per second instead of volume per second. The flow velocity may vary with the vessel's diameter, which is usually unknown (e.g., Lunt et al. 2004). Furthermore, the angle of the insonating beam with the flow direction of the artery may differ from one investigation to another. This, too, causes variation in FV measurements.
Regarding the physiologic factors, the middle cerebral artery flow velocity (MCAFV) depends on proximal influences such as heart rate, stroke volume, arterial blood pressure, possible obstruction in carotid arteries, on distal factors such as peripheral resistance, intracranial pressure etc., as well as on rheologic factors such as hematocrit and blood viscosity. All these factors, of which many are not readily known, should ideally be taken into account when interpreting the MCAFV signal (McCartney et al. 1997).
Many clinical TCD studies are conveniently based upon the parameters provided by commercial TCD apparatus (e.g., Hanlo et al. 1995; Rainov et al. 2000). We should, however, realize that this set of parameters has major theoretical drawbacks. These are summarized in Table 1. First, peak systolic flow velocity is the maximal flow velocity encountered anywhere during systole. TCD-manufacturers have made no distinction between early and late systole. This article will demonstrate that it is important to distinguish two phases in systole: stroke onset (or sys1) and the remaining part of systole (sys2). Second, the variation in insonation angle and vessel diameter result in large standard deviations in measurements of mean, diastolic and systolic flow velocities, which complicate interindividual comparison. This effect can be removed when all FV measurements are taken relative to a reference measurement of diastolic FV. Finally, for the calculation of a pulsatility index (PI), most commercial apparatus rely on a mean flow velocity and an end-diastolic flow velocity. By their definition, both measurements largely depend on heart rate. To avoid heart rate dependency, we propose that the reference diastolic flow velocity mentioned above should be taken at a fixed time interval with respect to stroke onset.
These theoretical considerations have led to the formulation of a new set of parameters. We tested this new set of parameters in a group of patients with significant internal carotid artery stenosis in comparison with a group of normal subjects without stenosis.
Section snippets
Materials and Methods
At our hospital, all patients selected for carotid surgery routinely undergo a preoperative assessment of intracranial hemodynamics by means of TCD. The rationale for this assessment is formed by the belief that patients with more severe hemodynamic disturbance from carotid stenosis have a greater risk upon postoperative hyperperfusion (Sbarigia et al. 1993; Reigel et al. 1987; Jorgensen and Schroeder 1993; Schaafsma et al. 2002). Hemodynamic status can be determined by measuring CO2-reactivity
Results
To better grasp the differences in FV wave morphology we determined values for acceleration, meaning the maximal steepness of the FV increase at stroke onset and for the first and second systolic peak normalized with respect to the dias@560 (see Methods section). These new parameters were compared with the parameters used traditionally: the mean and PI.
Table 2 contains data on the statistical analysis of all parameters when comparing the group with significant carotid stenosis to a group of
Discussion
As outlined in the Introduction section and summarized in Table 1, the current parameterization of the TCD signal suffers from important theoretical drawbacks. This is illustrated by Figure 3. The TCD parameters most often used, namely the mean and PI, are unable to distinguish a patient group with significant ipsilateral stenosis from a normal group, even though, in addition, there was a significant age difference between both groups. This article introduces a new set of parameters aimed to
Conclusion
The present data shows that the proposed set of TCD parameters allows for a good discrimination of carotid artery patients from normal controls. In the future, we aim to further test these parameters in health and disease and thereby increase our understanding of the interpretation of the TCD signal. Ultimately, we hope that TCD will evolve into the neurovascular monitoring tool it was anticipated, given its bedside availability and high temporal resolution.
Acknowledgments
The author gratefully acknowledges the support of Anuschka Niemeijer, Research Institute Martini Ziekenhuis Groningen, for statistical analyses and textual remarks and of Stefan van Duijvenboden and Bernard J. Geurts for help with the automatic analysis of sys1 and sys2 components.
The author is owner of MEAR Holding BV, an institution aiming to improve neuromonitoring on the ICU. MEAR Holding BV. holds an international patent on ‘PaR Technology’, based on a simultaneous analysis of arterial
References (18)
- et al.
Transcranial Doppler sonography pulsatility index (PI) reflects intracranial pressure (ICP)
Surg Neurol
(2004) - et al.
Defective cererbral autoregulation after carotid endarterectomy
Eur J Vasc Surg
(1993) - et al.
Cerebral hyperperfusion syndrome: A cause of neurologic dysfunction after carotid endarterectomy
J Vasc Surg
(1987) - et al.
Three cases of hyperperfusion syndrome identified by daily transcranial Doppler investigation after carotid surgery
Eur J Vasc Endovasc Surg
(2002) - et al.
Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries
J Neurosurg
(1982) - et al.
Neuromonitoring in the intensive care unit. 1. Intracranial pressure and cerebral blood flow monitoring
Intensive Care Med
(2007) - et al.
Wave intensity analysis from the common carotid artery: A new noninvasive index of cerebral vasomotor tone
Heart Vessels
(2003) - et al.
Differentiating vascular pathophysiological states by objective analysis of flow dynamics
J Neuroimaging
(2004) - et al.
Continuous assessment of cerebral autoregulation: Clinical and laboratory experience
Acta Neurochir Scand
(2003)
Cited by (10)
A Review of the use of Transcranial Doppler Waveform Morphology for Acute Stroke Assessment
2020, Journal of Clinical NeuroscienceA new method for correcting middle cerebral artery flow velocity for age by calculating Z-scores
2018, Journal of Neuroscience MethodsCitation Excerpt :In addition, the following information was retrieved from the medical record: age, gender and medication at the time of autonomic function testing. The so called envelope of the Doppler shifted signal, i.c. the maximum blood flow velocity encountered within the Doppler spectrum, was used to determine the TCD parameters (Schaafsma, 2012) using customised software developed at the Martini Ziekenhuis Groningen within a LabVIEW environment (National Instruments, version 7.1.1, USA): The ensemble average over all individual heartbeats within a 10 s period was used to determine the following parameters (Fig. 3): 1) acceleration (Acc), 2) first systolic peak (Sys1), 3) second systolic peak (Sys2), 4) diastolic flow velocity (Dias@560) and 5) HR.
Fluid Resuscitation in Septic Patients Improves Systolic but not Diastolic Middle Cerebral Artery Flow Velocity
2017, Ultrasound in Medicine and BiologyCitation Excerpt :On the basis of power analysis, we aimed to include a total of 16 patients in this study. Previous results from the study by Schaafsma (2012) were used for sample size calculation. A 2-sided sample size calculation with mean sys1-sys2 at the beginning of fluid resuscitation = −0.1, mean sys1-sys2 at the end of fluid resuscitation = 0.2, standard deviation (SD) = 0.2, α = 0.05 and power = 0.80 determined that we needed to include a total number of 16 patients.
Harvey with a modern twist: How and why conducting arteries amplify the pressure wave originating from the heart
2014, Medical HypothesesCitation Excerpt :This would allow us to discriminate between Sys1 and Sys2 hypertension, each of which is likely to require a tailored and distinct therapeutic approach. In relation to Transcranial Doppler measurement, the discrimination of Sys1 and Sys2 components has already improved the parameterization of the blood flow velocity signal [17]. It is not easy to predict findings in the future and for the moment the case will have to rest.