Slow Rhythmic Oscillations of Blood Pressure, Intracranial Pressure, Microcirculation, and Cerebral Oxygenation
Dynamic Interrelation and Time Course in Humans
Background and Purpose Various biological signals show nonpulsatile, slow rhythmic oscillations. These include arterial blood pressure (aBP), blood flow velocity in cerebral arteries, intracranial pressure (ICP), cerebral microflow, and cerebral tissue Po2. Generation and interrelations between these rhythmic fluctuations remained unclear. The aim of this study was to analyze whether stable dynamic interrelations in the low-frequency range exist between these different variables, and if they do, to analyze their exact time delay.
Methods In a clinical study, 16 comatose patients with either higher-grade subarachnoid hemorrhage or severe traumatic brain injury were examined. A multimodal digital data acquisition system was used to simultaneously monitor aBP, flow velocity in the middle cerebral artery (FVMCA), ICP, cerebral microflow, and oxygen saturation in the jugular bulb (Sjo2). Cross-correlation as a means to analyze time delay and correlation between two periodic signals was applied to a time series of 30 minutes' duration divided into four segments of 2048 data points (≈436 seconds) each. This resulted in four cross-correlations for each 30-minute time series. If the four cross-correlations were consistent and reproducible, averaging of the original cross-correlations was performed, resulting in a representative time delay and correlation for the complete 30-minute interval.
Results Reproducible cross-correlations and stable dynamic interrelations were found between aBP, FVMCA, ICP, and Sjo2. The mean time delay between aBP and ICP was 6.89±1.90 seconds, with a negative correlation in 81%. A mean time delay of 1.50±1.29 seconds (median, 0.85 seconds) was found between FVMCA and ICP, with a positive correlation in 94%. The mean delay between ICP and Sjo2 was 9.47±2.21 seconds, with a positive correlation in 77%. Mean values of aBP and ICP did not influence the time delay and dynamic interrelation between the different parameters.
Conclusions These results strongly support Rosner's theory that ICP B-waves are the autoregulatory response of spontaneous fluctuations of cerebral perfusion pressure. There is casuistic evidence that failure of autoregulation significantly modifies time delay and the correlation between aBP and ICP.
Rhythmic oscillations are a common phenomenon in biology, occurring on a metabolic level as well as in highly complex regulatory mechanisms (eg, glycolysis, endocrinological rhythms). Slow, non–heart beat–related rhythmic oscillations of ICP with a frequency of 0.5 to 2.0/min (B-waves) were first described by Lundberg,1 who attributed them to the Cheyne-Stokes respiration of his nonintubated patients with concomitant Pco2 fluctuations. A variety of different factors were found to be responsible and characteristic for a mean or static ICP increase, but although quite a number of different studies2 3 4 5 6 7 8 9 10 11 12 13 14 on the dynamic behavior of ICP waves have been performed since that time, the pathophysiological understanding of the dynamics of ICP waves and their interrelation with other systemic (aBP) and cerebral vascular or metabolic phenomena remains amazingly unclear.
Since B-waves are also observed in ventilated patients with a steady Pco2, other mechanisms must be responsible at least in part for the observed phenomena. Einhäupl et al3 and Venes13 postulated an autonomic (brain stem) rhythm as the direct pacemaker of ICP fluctuations with simultaneous (cardio)vascular oscillations. De Rougement et al9 and Yamamoto et al14 reported that aBP waves would follow ICP waves with a latency of 5 to 10 seconds and thus concluded that ICP fluctuations would induce an instability of the vasomotor center followed by blood pressure oscillations. Magnæs4 was one of the first to postulate that B-waves were generated by changes in intracranial blood volume reflecting blood pressure waves and brain autoregulation.
In 1983 Rosner and Becker10 postulated that fluctuations of CPP evoke an autoregulatory response of the cerebral vessels leading to ICP oscillations due to fluctuations of intracranial volume. The correlation and exact temporal relation between these phenomena remained more or less unknown.
Rhythmic slow fluctuations of the same frequency range were also observed to occur in cerebral FV measured by transcranial Doppler sonography5 6 and cerebral hemoglobin oxygenation in the jugular bulb (Sjo2).15 For a long time, rhythmic slow waves were also observed in the brain tissue Po2, the cortical cytochrome c oxidase, and the cortical blood volume as measured by reflectance spectroscopy. The cerebral microvasculature shows slow rhythmic phenomena known as vasomotion. It remains unclear whether these different microcirculatory and metabolic oscillations are local or regional phenomena to supply local metabolic demands, are synchronized by unknown mechanisms, or are induced by systemic hemodynamic factors.
Different approaches were undertaken to understand the interrelation between ICP waves and other parameters. Most groups analyzed the waveform of the heart beat–dependent pulsatile fluctuation of ICP, ie, the propagation of the pulse wave from the arterial bed to the intracranial space.7 8 16 17 18 Piper et al8 analyzed phase shift and gain by a transfer-function analysis of the heart beat–related fluctuations between aBP and ICP and concluded from their experimental work that a loss of autoregulation modifies the phase shift between aBP and ICP. Only Giller19 applied time series analysis to analyze the low-frequency, non–heart beat–related oscillations of aBP and FVMCA during neurosurgical procedures for coherence and gain. Giller interpreted the occurrence of high correlations (coherence) between aBP and FVMCA as an indicator for impaired autoregulation.
Thus, the dynamic interrelations between the low-frequency oscillations of aBP, FVMCA, ICP, and microvascular and metabolic cerebral phenomena have not yet been analyzed thoroughly by means of time series analysis.
The aim of this study was to analyze whether stable dynamic interrelations in the low-frequency range exist between these different variables, and if they do, to analyze their exact time delay. In addition, a correlation analysis between the different static mean values and their influence on the time delay of the dynamic interrelations was performed.
Subjects and Methods
Sixteen individuals (11 men and 5 women) were investigated (mean age, 48.1 years; SD, 16.2; median, 47 years). Ten patients suffered from severe SAH (Hunt and Hess grade 4 or 5), and six had severe TBI (Glasgow Coma Scale score <7).
Procedures and Monitoring Protocol
Intraventricular drainage and two subdural laser Doppler disk probes were inserted through two separate frontal burr holes in the operating theater after admission. If necessary, evacuation of an intracranial hematoma was done preferentially. Laser Doppler probes were placed on the cortical surface of the frontal lobe. Additional probes (for aBP, FVMCA, Sjo2, and end-tidal Pco2 monitoring) were inserted or applied in the intensive care unit. Monitoring in the intensive care unit according to the protocol started as early as possible after admission, which was approximately 6 to 8 hours after admission because of the time necessary for probe implantation, cabling, and routine calibration procedures.
The monitoring protocol (as approved by the Ethical Committee of the University of Erlangen-Nürnberg, No. 435/94) consisted of three recording sessions of 30 minutes every 8 hours on each of 3 consecutive days. Patients were artificially ventilated, and Paco2 was maintained at 31±2 mm Hg. During data acquisition, any manipulation of the patient (eg, nursing, administration of drugs) was avoided. If necessary, patients were sedated with 5 to 20 mg of midazolam given 5 to 10 minutes before monitoring. At least 30 minutes before and during the entire recording period, ventricular drainage was kept closed. For ethical reasons, ventricular drainage was opened and/or aBP was adjusted only if ICP exceeded 25 mm Hg or CPP decreased below 60 mm Hg, and the complete data set was excluded from further analysis.
Biological Signals Measured
aBP, ICP, FVMCA, cortical microflow, Sjo2, systemic Sao2, and end-tidal Pco2 were monitored continuously and simultaneously.
A catheter tip pressure transducer (type 110-4HM, Camino 420 system, Camino Laboratories) was used to monitor aBP and ICP. Catheters were inserted into the femoral artery and into the lateral cerebral ventricle. FVMCA was measured by a 2-MHz pulsed transcranial Doppler device (Trans-scan, EME). The temporal window of the ipsilateral side of ICP monitoring was used, and the skin at the optimal position was marked for repetitive measurements. The ultrasound probe was fixed to the patient's head with an elastic strap. Microflow was monitored with two probes (disk probes, type PD434) by means of laser Doppler flowmetry (Laserflo, BPM 403 A, Vasamedics). The velocity signal (range, 0 to 8 kHz; “Veloc”), the fractional volume of the red blood cells (range, 0 to 1.6 arbitrary units; “Vol”), and microflow (scaled product of velocity and volume) (range, 0 to 400 arbitrary units; “Flux”) plus the light intensity of the backscattered light were measured simultaneously. Measurement of Sjo2 was performed with the Oximetrix 3 System (Abbott Laboratories) after in vitro calibration, retrograde cannulation of the internal jugular vein, and placement of the fiberoptic catheter (type Opticath U425) within the jugular bulb. Catheter tip position was monitored by a daily x-ray film, and in vivo calibration was also performed daily. Because of the sampling technique of the Oximetrics device (the actual value is calculated as the moving average of the last 5 seconds, and a new sample is added every second), its mean inherent time delay is approximately 2 seconds, depending on the main frequency of Sjo2 fluctuations.
All analog data were digitized by two 16-channel, 12-bit resolution analog-to-digital convertors (type DT2814, Data Translation, Inc), with an analog-to-digital convertor throughput of 25 kHz. A specially designed software program for a personal computer was used, which was written by U. Sigwanz using the TopSpeed environment with Modula-2 (Jensen & Partners UK Ltd). The program allows free control of the sampling rate and duration of the sampling interval. Since the nonpulsatile, non–heart cycle–dependent pattern of the different signals was the target of analysis, the sampling rate was set to 4 per second and an interval of 30 minutes stored for later off-line analysis.
To exclude aliasing effects, analog output signals were filtered before digitization by an analog low-pass filter (3-dB cutoff frequency at 50 Hz). To remove heart cycle–dependent pulsatility, an additional low-pass filter with a 3-dB cutoff frequency at 0.1 Hz was applied before data were stored. Thus, frequencies below 0.1 Hz (≈6 cycles per minute) were taken for further analysis.
Methods for Time Series Analysis
Cross-correlation function is appropriate for detecting (1) the time delay and (2) the correlation between periodic oscillations of two signals.20 The method is highly reliable if periodic input signals share the same main frequency (Fig 1⇓).
Cross-correlation may lead to misinterpretation concerning physiological data because of their spectral complexity.20 Therefore, calculations were done on four sequential time windows, each consisting of 2048 data points per signal, equivalent to a time interval of 436 seconds. Mean time delays of the different curves were averaged only if the cross-correlations of at least three consecutive intervals did not differ concerning correlation and not differ significantly concerning time delay (Fig 2⇓). Time series that did not have these characteristics were rejected from further analysis. This type of analysis yields unequivocal results, and the averaged time delay was therefore regarded as representative of the complete 30-minute recording and taken for further analysis.
The complete off-line analysis software was written by U. Sigwanz using the TopSpeed environment with Modula-2 (Jensen & Partners UK Ltd).
Cross-correlation of the dedicated time series X(t) and Y(t), denoted as cross-correlation of X→Y, was calculated from the Fourier coefficients with standard algorithms.21 22 Cross-correlations were performed for aBP→ICP, FVMCA→ICP, ICP→Sjo2, FVMCA→Sjo2, and Flux 1/Flux 2→ICP. We reported only the direction (positive/negative) of the correlations rather than absolute values (from −1 to +1).
We included 143 monitoring sessions from 16 patients for statistical analysis. Periods with artifacts (caused, for example, by displacement of the transcranial Doppler probe or low light intensity of the jugular venous fiberoptic catheter) were excluded from analysis.
Mean values of the complete 30-minute interval were obtained for each physiological parameter from all recording periods and denoted as “static” values. Analysis for normal distribution of mean values was done by the Kolmogorov-Smirnov test. Correlation between variables was tested by Spearman's rank correlation analysis.
Cross-correlations between aBP, ICP, FVMCA, laser Doppler flowmetry Flux, and Sjo2 were calculated as described above. The frequency distribution of the time delay was also calculated. Mean values of time delay were calculated for the cross-correlation results of aBP→ICP, ICP→FVMCA, ICP→Sjo2, and FVMCA→Sjo2.
A value of P<.05 was considered significant.
The statistical software package Statgraphics Plus for Windows (Manugistics, Inc) was used for statistical analysis.
Static Values of 30-Minute Recording Periods
Table 1⇓ shows static mean values for the different signals; no statistically significant differences were found between the two patient groups.
The ICP values in all patients were either normal or only slightly elevated because of the study protocol to open the ventricular drainage if ICP exceeded 25 mm Hg and to exclude these recording periods from further analysis. aBP, CPP, and Sao2 were within normal physiological limits (Table 1⇑), since in this patient group no major cardiovascular or pulmonary problems occurred.
Dynamic Interrelation: Time Delay and Correlation Between Signals
Reproducible cross-correlations were found in 80 of 122 time series concerning aBP→ICP, 102 of 108 series concerning FVMCA→ICP, 53 of 104 series concerning FVMCA→Sjo2, and 61 of 114 series concerning ICP→Sjo2 (the first number represents the number of reproducible cross-correlations; the second is the total number of cross-correlations performed). Cross-correlations performed between laser Doppler flowmetry Flux and ICP yielded inconsistent results in more than 50%. Thus, for further analysis only the four cross-correlations (aBP→ICP, FVMCA→ICP, FVMCA→Sjo2, and ICP→Sjo2) were taken into consideration.
The statistical analyses are summarized in Table 2⇓. The mean time delay between aBP and ICP was 6.89±1.90 seconds, and the correlation was negative in 65 of 80 analyses (81%). In a few (15 of 80) series, a positive correlation with significantly shorter time delay (mean, 0.62 second) was observed. In nearly all series (96 of 102; ≈94%), a positive correlation was found between FVMCA and ICP, and the mean time delay was found to be 1.50±1.29 seconds. A constant positive correlation was found between ICP and Sjo2 in 47 of 61 analyses (77%), with a mean time delay of 9.47±2.21 seconds. Accordingly, the time delay between FVMCA and Sjo2 was found to be 11.25±1.54 seconds, with a positive correlation.
The majority (11 of 15) of cross-correlations with positive correlation and short (<2 seconds) time delay concerning aBP→ ICP were observed in a single patient with severely disturbed autoregulation (see “Case Report”).
Correlation analysis between mean time delay shows a significant positive correlation of mean time delays for FVMCA→ICP and ICP→Sjo2 (0.51; P<.01) as well as for ICP→Sjo2 and FVMCA→Sjo2 (0.69; P<.01).
No correlations were found between the mean time delay of aBP→ICP, FVMCA→ICP, ICP→Sjo2, FVMCA→Sjo2 and the static mean values of aBP, FVMCA, CPP, ICP, and Sjo2.
A 74-year-old comatose woman was admitted to the department, and CT scans showed severe subarachnoid bleeding with ventricular hemorrhage, clinically grade 5 (Hunt and Hess). Angiography was not performed because of the poor condition of the patient. Multimodal monitoring was performed in the intensive care unit from day 1 until day 4 after bleeding. Results of the cross-correlations aBP→ICP during the first 3 days showed a positive correlation (Fig 3⇓, left panel). Concomitant autoregulation tests (with induced hypotension) showed a failure of autoregulatory response, evident by a reduction of Sjo2, decrease of mean FVs, and lack of ICP increase.
On day 4 after SAH, a negative correlation and time delay of 7.89 seconds between aBP and ICP waves were observed (Fig 3⇑, right panel). An autoregulation test performed on the same day showed a reestablished autoregulatory capacity. Induced hypotension now led to an increase of ICP, constant FVs, and even an increase of Sjo2. Nevertheless, the poor neurological status of the patient continued, and she died on the 16th day after bleeding.
Friesen and Block23 argued that for the analysis of biological oscillators, three questions are to be addressed: (1) What are the biological variables essential to the oscillator? (2) How do these essential variables interact? (3) Can these interactions lead to oscillations? They hypothesized that for any system with rhythmic behavior, several elements must be present: (1) an excitatory drive; (2) a restorative process turning the oscillating system toward steady state; and (3) an overshoot, which is generally produced by a time delay between the elements. Thus, in developing an approach to understand the interrelation between different variables showing rhythmic oscillations, the analysis of the factors involved, their correlation, and the time delay between them are essential factors.
An important question is whether a significant time delay is added by the measurement device itself. Since a low-pass filter is applied to all signals to remove heart beat–related phenomena, a significant time delay is added neither by the Camino tip catheters for pressure monitoring (ICP and aBP), nor by laser Doppler flowmetry (with a time constant of 0.2 second), nor by transcranial Doppler sonography (time resolution, 6 microsecond). Only the Oximetrics signals have an inherent time delay of approximately 2 seconds.
Analysis of the static values did not reveal any correlation between aBP, ICP, FVMCA, microflow, and Sjo2. An explanation for the missing correlation between ICP and Sjo2 values is the relatively “normal” ICP values due to our protocol without any massive ICP increase; in the case of ICP increase, the external ventricular drainage was opened and the recording period was excluded from further analysis. Furthermore, no correlation between the different static mean values and the time delay between the different variables was observed, which could be attributed to the normal or nearly normal static values.
In contrast, analysis of the dynamic behavior between the different parameters (Table 2⇑) clearly showed an inherent, complex interrelation between them. Although the analyses were performed in a heterogeneous patient group with both SAH and TBI, it seems most unlikely that differences in pathogenetic pathways between both entities influence the pathophysiological phenomena examined in this study. This is corroborated by a lack of difference in the variables analyzed between the two patient groups.
Different slow rhythmic aBP waves can be discriminated24 : either Hering-Traube waves (attributed to and closely related to respiratory activation of the brain stem) or Mayer waves (with a considerably lower frequency than the respiration rate, generated in the ventrolateral medullary surface). Ever since Rosner's theory10 11 12 of ICP A- and B-waves, CPP fluctuations play a central role in the discussion of the generation of ICP waves. Whereas Rosner's theory implies that ICP waves follow aBP oscillations, some authors9 14 have a contradictory opinion and presented observations showing that aBP waves followed ICP fluctuations. No time series analyses were performed until now to validate the very important temporal relation between fluctuations of aBP and ICP in the low-frequency range.
Our data clearly reveal a stable interrelation between aBP and ICP found in more than 60% of the series analyzed: In approximately 81%, ICP lagged significantly behind aBP waves (mean, 6.89 seconds), and the correlation was negative. This is in full agreement with Rosner's theory of the generation of ICP B-waves in cases of preserved autoregulation. Nevertheless, in 19% of the time series, a positive correlation with a very short time delay (mean, 0.62 second) was found. Two factors may be responsible for the time delay between aBP and ICP: (1) an active autoregulatory process by the cerebral vasculature and (2) the time for pulse pressure propagation of the blood pressure waves into the cerebral microvasculature, leading to ICP waves. Circulation time does not play a significant part, since in this case the time delay between pressure waves is of importance.
Aaslid et al25 studied the autoregulatory response in normal volunteers with different Pco2 levels using transcranial Doppler sonography and found a restoration of blood flow to the pretest level after 4.1 seconds under hypocapnia. An elevation of Pco2 increases the time necessary for FV restoration up to approximately 10 seconds. Discussing a study by Symon et al,26 Aaslid et al25 attached an initial autoregulation peak at 0.5 second to the transmission of flow waves to the vein of Labbé and the secondary peak at 3 to 4 seconds to the real autoregulatory response. Kontos et al27 studied responses of the vessel diameter in anesthetized cats and found a mean response time of 5.2 seconds depending on the rate of blood pressure reduction. Induced blood pressure waves generated oscillating changes of vessel diameter with negative phase and a time lag of approximately 10 seconds (Fig 10). Florence and Seylaz28 measured the autoregulatory readjustment to prehypotensive values using laser Doppler flowmetry in anesthetized rabbits and found autoregulation times between 2.8 and 13.2 seconds, depending on the degree of hypotension.
Two facts strongly support the theory that autoregulatory phenomena are responsible for the negative correlation and the time delay of 6.89 seconds between aBP and ICP. First, a negative correlation between two signals implies that an active process is responsible. If an active phenomenon would not interfere, a positive correlation between the signals is to be presumed. The second fact is the zero time delay in cases of positive correlation. This strongly excludes the possibility that the mean time delay of 6.89 seconds in cases of negative correlation is exclusively due to the time of pulse pressure propagation. In addition, the mean time for pulse pressure propagation of approximately 10 m/s29 excludes the possibility that passive physical phenomena are responsible. Since aBP was measured in the femoral artery, it can be presumed that central blood pressure fluctuations were registered predominantly.
For a long time, different oscillatory phenomena were also observed in various cerebral parameters other than aBP and ICP. Clark et al30 were the first to describe cyclic fluctuations of brain tissue Po2 measured with polarographic electrodes in the cortex of unanesthetized cats. Po2 waves were observed with a frequency of 6 to 12/min and were interpreted to be local phenomena, since distantly located probes revealed Po2 fluctuations asynchronously, and the fluctuations showed no correlation with respiration, heart rate, or blood pressure. These “oxygen availability waves” were also observed in humans.31 Halsey and McFarland32 observed that the amplitude of tissue Po2 oscillations with a frequency of 3 to 8/min depended on mean aBP (increasing amplitude with decreasing aBP).
Cycling of cortical cerebral blood flow was described by several groups. Auer and Gallhofer35 observed rhythmic diameter oscillations of cat pial vessels within a wide range of frequencies between 0.5 and 8 cycles per minute. Hundley et al36 described lower frequencies at 0.74 cycles per minute of vessel diameter in unanesthetized rabbits. Cycling of cerebral blood flow was also observed by laser Doppler flowmetry (“fluxmotion”) in cerebral tissue by Morita-Tsuzuki et al37 (8 to 10 cycles per minute, in rats), Dirnagl et al38 (0.5 to 8 cycles per minute, in rats), and Jones et al39 (mean of 0.094 Hz, in rats). The different frequencies of aBP waves and ICP waves on the one hand and fluxmotion on the other hand seem to be responsible for the unstable cross-correlations between these parameters in the present study.
Mayevsky and Ziv40 performed an extensive study in rats to examine the phase relation between “spontaneous” oscillations of NADH and flavoprotein as well as tissue Po2 and laser Doppler flowmetry flux and volume. With the use of simultaneous recordings from different locations on the brain surface, it was obvious that a local and not a systemic generator for the oscillation of the different parameters must be postulated, since phase and frequency differed between different locations. The authors postulated that vasomotion is the initial generator of the dependent Po2 and NADH fluctuations.
Non–heart beat–related, slow fluctuations of FV measured by transcranial Doppler sonography were observed by several authors5 6 41 and were found to be phase-locked and in-phase with ICP fluctuations.6 Auer and Sayama2 observed rhythmic fluctuations of the pial vessel wall diameter simultaneously with B-waves and postulated that B-waves were a sum phenomenon of intracranial volume oscillations mediated by pial vessels. Newell et al6 performed simultaneous recordings of FVMCA on both sides with two transcranial Doppler devices and found corresponding rhythmic fluctuations bilaterally, indicating that one individual FVMCA signal is characteristic of the global cerebral vasculature. Our findings support the observations of Newell et al of a constant positive correlation between FVMCA and ICP.
Although some authors6 19 were interested in the correlation between blood flow velocity and ICP, this time delay is a complex phenomenon to analyze, since the propagation of corpuscles (FVMCA) is compared with the propagation of pressure waves (aBP, ICP). The propagation velocity of both differs significantly. Whereas pulse pressure propagation is approximately 10 m/s,29 circulation time is a much slower process. Okawara et al42 found the angiographically determined mean circulation time (from the carotid siphon to the large cortical parietal veins) in patients with SAH to be much longer (7.2 seconds) than normal (5.4±0.53 seconds).
The time delay of 9.47 seconds (device-related delay, 2 seconds) plus the positive correlation between ICP and Sjo2 or 11.25 seconds between FVMCA and Sjo2 is presumed to be the net circulation time of oxygenated blood from the middle cerebral artery through the cerebral capillaries to the jugular bulb. Waves of oxygen inflow are obviously transmitted through the brain into the jugular bulb during constant cerebral metabolism. Whereas vasomotion and brain tissue Po2 are obviously local phenomena,40 cycling of hemoglobin oxygenation in the jugular bulb is strongly related to general oxygen “income” cycling. The pulse pressure propagation of aBP→ICP should be interpreted as a phenomenon separate from FVMCA→Sjo2 (circulation).
Cross-correlation as applied in this series is a valid tool to analyze the complex dynamic behavior of hemodynamic and metabolic cerebral parameters by evaluating the time delay plus the correlation between the different signals.
The Cambridge group43 44 recently reported on possibilities for continuous autoregulation monitoring using the “floating or moving correlation coefficient” between diastolic FVMCA or amplitude of FVMCA and CPP. We speculate that autoregulatory mechanisms are responsible for the time delay between aBP and ICP plus the negative correlation between them, and we presume that a failure to autoregulate manifests in a zero or nearly zero time delay and a positive correlation between the signals. If the theory holds true, this phenomenon might have a practical, clinical consequence, since continuous autoregulation monitoring by means of continuous cross-correlation analysis may become a possibility in the near future. The examination of autoregulatory capacity was not the original aim of this study; therefore, further experimental and prospective clinical studies should be done to test the applicability, sensitivity, and clinical robustness of continuous cross-correlation autoregulation monitoring.
Selected Abbreviations and Acronyms
|aBP||=||arterial blood pressure|
|CPP||=||cerebral perfusion pressure|
|FV||=||blood flow velocity|
|FVMCA||=||blood flow velocity in the middle cerebral artery|
|Sao2||=||arterial oxygen saturation|
|Sjo2||=||oxygen saturation in the jugular bulb|
|TBI||=||traumatic brain injury|
This study was supported by the Deutsche Forschungsgemeinschaft, Bonn, Germany (Ste 512/1-2). The software for digital data acquisition was written by Ullrich Sigwanz. The authors appreciate his thorough knowledge, understanding, and constant help during establishment of the multimodal data acquisition system. The authors thank Andreas Dötterl, Markus Kleemann, and Dierk Ronneberger for their invaluable help during data acquisition. Statistical methods were reviewed by P. Martus, Institut für Medizinische Dokumentation und Statistik, Universität Erlangen-Nürnberg, Erlangen. Many thanks to Andreas Raabe, Chemnitz, Germany, and Marek Czosnyka, Cambridge, UK, for supportive discussions concerning possibilities of continuous autoregulation monitoring.
Reprint requests to Ralf Steinmeier, MD, Neurochirurgische Klinik, Universität Erlangen-Nürnberg, Schwabachanlage 6, 91054 Erlangen, Germany. E-mail email@example.com.
- Received March 29, 1996.
- Revision received August 23, 1996.
- Accepted August 25, 1996.
- Copyright © 1996 by American Heart Association
Lundberg N. Continuous recording and control of ventricular fluid pressure in neurosurgical practice. Acta Psych Neurol Scand Suppl. 1960;36:1-193.
Einhäupl KM, Garner C, Dirnagl U, Schmieder G, Schmiedek P, Kufner G, Rieder J. Oscillations of ICP related to cardiovascular parameters. In: Miller JD, Teasdale GM, Rowan JO, eds. Intracranial Pressure VI. Berlin, Germany: Springer-Verlag; 1986:290-297.
de Rougement J, Barge M, Benabid AL. Correlative study of intracranial and systemic arterial pressures in neurosurgical patients. In: Brock M, Dietz H, eds. Intracranial Pressure I. Berlin, Germany: Springer-Verlag; 1972:232-238.
Rosner MJ, Becker DP. The etiology of plateau waves: a theoretical model and experimental observations. In: Ishii S, Nagai H, Brock M, eds. Intracranial Pressure V. Berlin, Germany: Springer-Verlag; 1983:301-306.
Rosner MJ. The vasodilatory cascade and intracranial pressure. In: Miller JD, Teasdale GM, Rowan JO, Galbraith SL, Mendelow AD, eds. Intracranial Pressure VI. Berlin, Germany: Springer-Verlag; 1986:137-141.
Yamamoto S, Higashi S, Fujii H, Hayashi M, Ito H. Vasomotor response in acute intracranial hypertension. In: Ishii S, Nagai H, Brock M, eds. Intracranial Pressure V. Berlin, Germany: Springer-Verlag; 1983:333-337.
Portnoy HD, Chopp M, Branch C, Shannon MB. Cerebrospinal fluid pulse waveform as an indicator of cerebral autoregulation. J Neurosurg. 1982;56:668-678.
Stearns SD, David RA. Signal Processing Algorithms. Englewood Cliffs, NJ: Prentice-Hall; 1988.
Press WH, Flannery BP, Teukolsky SA, Vetterling WT. Numerical Recipes in C. Cambridge, UK: Cambridge University Press; 1988.
Friesen WO, Block GD. What is a biological oscillator? Am J Physiol. 1984;246(Regulatory Integrative Comp Physiol. 15):R847-R851.
Haxhiu M, van Lunteren E, Deal E, Cherniack N. Role of the ventral surface of medulla in the generation of Mayer waves. Am J Physiol. 1989;257(Regulatory Integrative Comp Physiol. 26):R804-R809.
Aaslid R, Lindegaard K, Sorteberg W, Nornes H. Cerebral autoregulation dynamics in humans. Stroke. 1989;20:45-52.
Symon L, Held K, Dorsch NWC. A study of regional autoregulation in the cerebral circulation to increased perfusion pressure in normocapnia and hypercapnia. Stroke. 1973;4:139-147.
Kontos HA, Wei PW, Navari RM, Levasseur JE, Rosenblum WI, Patterson JL Jr. Responses of cerebral arteries and arterioles to acute hypotension and hypertension. Am J Physiol. 1978;234(Heart Circ Physiol. 3):H371-H383.
Busse R. Kreislaufphysiologie. Stuttgart, Germany: G. Thieme Verlag; 1982.
Clark LC, Misrahy G, Fox RP. Chronically implanted polarographic probes. J Appl Physiol. 1958;13:85-91.
Gretchin VB. Some data on oxygen dynamics in subcortical structures of human brain. Electroencephalogr Clin Neurophysiol. 1969;26:546-547.
Halsey JH, McFarland S. Oxygen cycles and metabolic autoregulation. Stroke. 1974;5:219-225.
Hundley WG, Renaldo GJ, Levasseur JE, Kontos HA. Vasomotion in cerebral microcirculation of awake rabbits. Am J Physiol. 1988;254(Heart Circ Physiol. 23):H67-H71.
Jones SJ, Williams JL, Shea M, Easley KA, Wie D. Cortical cerebral blood flow cycling: anesthesia and arterial blood pressure. Am J Physiol. 1995;268(Heart Circ Physiol. 37):H569-H575.
Mayevsky A, Ziv I. Oscillations of cortical oxidative metabolism and microcirculation in the ischemic brain. Neurol Rev. 1991;13:39-47.
Czosnyka M, Guazzo E, Iyer V, Kirkpatrick P, Smielewski P, Whitehouse H, Pickard JD. Testing of cerebral autoregulation in head injury by waveform analysis of blood flow velocity and cerebral perfusion pressure. Acta Neurochir (Wien). 1994;60(suppl):468-471.