Secondary Logo

Journal Logo

The Limitations of Near-Infrared Spectroscopy to Assess Cerebrovascular Reactivity: The Role of Slow Frequency Oscillations

Diedler, Jennifer MD*,†; Zweifel, Christian MD*; Budohoski, Karol P. MD*; Kasprowicz, Magdalena PhD*; Sorrentino, Enrico MD*; Haubrich, Christina MD*; Brady, Kenneth M. MD*,‡; Czosnyka, Marek PhD*; Pickard, John D. FMedSci*; Smielewski, Peter PhD*

doi: 10.1213/ANE.0b013e3182285dc0
Neuroscience in Anesthesiology and Perioperative Medicine: Research Reports
Free
SDC

BACKGROUND: A total hemoglobin reactivity index (THx) derived from near-infrared spectroscopy (NIRS) has recently been introduced to assess cerebrovascular reactivity noninvasively. Analogously to the pressure reactivity index (PRx), THx is calculated as correlation coefficient with arterial blood pressure (ABP). However, the reliability of THx in the injured brain is uncertain. Although slow oscillations have been described in NIRS signals, their significance for assessment of autoregulation remains unclear. In the current study, we investigated the role of slow oscillations of total hemoglobin for NIRS-based cerebrovascular reactivity monitoring.

METHODS: This study was based on a retrospective analysis of data that were consecutively recorded for a different project published previously. Thirty-seven patients with traumatic brain injury and admitted to Addenbrooke's Neurosciences Critical Care Unit between June 2008 and June 2009 were included. After artifact removal, we performed spectral analysis of the tissue hemoglobin index (THI, a measure of oxy- and deoxygenated hemoglobin) and intracranial pressure (ICP) signal. PRx and THx were calculated as moving correlations between ICP and ABP, and THI and ABP, respectively. The agreement between PRx and THx as a function of normalized power of slow oscillations (0.015–0.055 Hz) contained in the input signals was assessed performing between-subject and within-subject correlation analyses. Furthermore, the correlation between the THx values derived from the right and left sides was analyzed.

RESULTS: The agreement between PRx and THx depended on the power of slow oscillations in the input signals. Between-subject comparisons revealed a significant correlation between THx and PRx (r = 0.80, 95% confidence interval 0.53–0.92, P < 0.01) for patients with normalized slow wave activity >0.4 in the THI signal, compared with r = 0.07 (95% confidence interval −0.40 to 0.51, P = 0.79) in the remaining files. Furthermore, within-subject comparisons suggested that THx may be used as a substitute for PRx only when there is an at least moderate agreement (r = 0.36) between the THx values derived from the right and left sides.

CONCLUSIONS: Our results suggest that the NIRS-based cerebrovascular reactivity index THx can be used as a noninvasive substitute for PRx, but only during phases with sufficient slow wave power in the input signal. Furthermore, a good agreement between the THx measures on both sides seems to be a prerequisite for comparison of a global (PRx) versus the more local (THx) index. Nevertheless, noninvasive assessment of cerebrovascular reactivity may be desirable in patients without ICP monitoring and help to guide ABP management in these patients.

Published ahead of print August 4, 2011 Supplemental Digital Content is available in the text.

From the *Academic Neurosurgical Unit, University of Cambridge Clinical School, Cambridge, United Kingdom; Department of Neurology, University of Heidelberg, Germany; and Department of Anesthesiology and Pediatrics, Texas Children's Hospital, Baylor College of Medicine, Houston, Texas.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.anesthesia-analgesia.org).

Study funding information is provided at the end of the article.

This report was previously presented, in part, at the ICP Conference 2010, Tübingen.

Conflict of Interest: See Disclosures at the end of the article.

Authors' current affiliations are listed at the end of the article.

Reprints will not be available from the authors.

Address correspondence to Jennifer Diedler, MD, Department of Neurology, University of Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany. Address e-mail to jennifer.diedler@med.uni-heidelberg.de.

Accepted May 17, 2011

Published ahead of print August 4, 2011

Cerebral autoregulation can be assessed noninvasively using near-infrared spectroscopy (NIRS).13 Recently, the NIRS-derived tissue hemoglobin index (THI) has been used to establish the cerebrovascular reactivity index THx (total hemoglobin reactivity index).4,5 THI is a normalized measure of relative oxy- and deoxygenated hemoglobin concentration and thereby provides a tracer of cerebral blood volume (CBV). THx is hypothetically equivalent to the already established pressure reactivity index (PRx) that is calculated as a moving correlation between intracranial pressure (ICP) and arterial blood pressure (ABP).6,7 Analogously, THx is calculated as the moving correlation between THI and ABP. Both indices aim to assess cerebrovascular pressure reactivity, defined as the ability of the cerebrovascular smooth muscles to react to changes in transmural pressure.6 Assessment of PRx and THx relies on the assumption that slow oscillations of ICP, and THI respectively, mirror fluctuations of CBV.8 When the cerebral vessels are reactive, a change in ABP will produce an inverse change in CBV provoking parallel changes in ICP or THI (negative correlation). In contrast, when cerebrovascular pressure reactivity is disturbed, changes in ABP are passively transmitted to changes of CBV and thereby to changes in ICP or THI (positive correlation).

In a piglet model, high coherences between ICP and THI were found in the slow wave frequency spectrum (0.005–0.04 Hz).5 This finding has been postulated to be the theoretical prerequisite and conditio sine qua non for NIRS-based assessment of cerebrovascular reactivity. Furthermore, it was shown that the hemoglobin volume index (theoretically equivalent to THx, but using another NIRS device with a distinct algorithm for calculation of the blood volume index), significantly correlated with PRx (r = 0.73).5 THx has been validated recently against PRx in patients with traumatic brain injury (TBI).4 A significant agreement between THx and PRx was found (correlation r = 0.56, P < 0.001), leading to the suggestion that THx may be used to optimize a cerebrovascular-reactivity–oriented therapy7 in patients for whom ICP monitoring is not feasible. However, it has been indicated that assessment of THx may not be reliable and applicable in all clinical situations, for instance, in patients with frontal contusions, that negatively affect NIRS recordings. In some patients, however, no explanation for failure of NIRS-based assessment of cerebrovascular reactivity was found. Furthermore, it remains unclear whether THx, based on a locally measured variable, the THI signals on the left and right sides, is suitable to draw conclusions on the global state of cerebrovascular reactivity.

Based on a data pool drawn from the same population as the aforementioned study,4 we aimed to investigate whether waveform analysis of the input signals is helpful to identify situations in which the NIRS-derived variable THx can be reliably used as a noninvasive correlate of PRx. With respect to the above-mentioned pathophysiological assumptions, we assessed the hypothesis that the presence of slow waves in the input signals could estimate the reliability of THx as an index of cerebrovascular reactivity. Furthermore, a comparison between the THx derived from the right and left THI signals was performed. A reliable noninvasive index to estimate cerebrovascular reactivity would be of high value to guide ABP management in patients in whom invasive ICP monitoring is not applicable.

Back to Top | Article Outline

METHODS

Patients

Data from 37 patients with TBI were included in the current analysis. All patients were admitted to Addenbrooke's Neurosciences Critical Care Unit between June 2008 and June 2009. Data were acquired as part of a research project investigating cerebral physiology and metabolism after TBI, approved by the local ethics committee. The data used in the current study have been analyzed and published, focusing on a different question.4 Written informed assent was obtained from the next of kin of each patient. Inclusion criteria were age older than 16 years, closed head injury, availability of ICP monitoring, and informed assent.

The patients were managed according to Addenbrooke's Neurosciences Critical Care Unit protocol.9 All patients' lungs were mechanically ventilated, and propofol and fentanyl were used for sedation. According to the inclusion criteria, all had ICP and cerebral perfusion pressure (CPP) monitoring. ABP was measured at the heart level without correction of CPP for the distance from heart to tragus. CPP was maintained above 60 mm Hg and ICP below 20 mm Hg. Temperature was kept at ≤37°C. Ventilator settings were adjusted to keep arterial oxygen pressure ≥84 mm Hg and arterial carbon dioxide pressure at approximately 34 mm Hg. Baseline clinical data were collected and neurological status on admission was graded according to the Glasgow coma scale score. Computed tomographic (CT) scans were graded according to the modified Marshall CT scan classification (for details on scoring, see Table 1).10 Frontal contusions were defined as intraparenchymal hyperdensity surrounded by hypodensity on the CT scan. Outcome at 3 months was assessed using the Glasgow outcome scale score (1 = dead, 2 = vegetative state, 3 = severe disability, 4 = moderate disability, and 5 = good recovery).

Table 1

Table 1

Back to Top | Article Outline

Signal Acquisition

The NIRO-200 monitor (Hamamatsu Photonics UK Ltd., Hertfordshire, UK) was used to monitor cerebral tissue oxygenation and cerebral total hemoglobin. The NIRO-200 generates 3 wavelengths of infrared light (775, 810, and 850 nm) using 1 emitting laser diode and 2 detecting photodiodes to measure light attenuation at different frequencies and different spatial separation from the source. Using a mathematical model, based on the light diffusion equations rather than the modified Beer-Lambert law, uncalibrated concentrations of oxygenated (CO2Hb) and deoxygenated hemoglobin (CHb) are calculated. Derived variables include the tissue oxygenation index (percentage value of CO2Hb/[CO2Hb + CHb]), and the THI (the sum of CO2Hb and CHb). It has been shown that these indices are not affected by the extracranial circulation and that they are independent of hemoglobin concentration, skull thickness, and the area of cerebrospinal fluid beneath the optodes.11,12 The NIRS THI signal was measured bilaterally over the frontal areas and digitally transferred to the recording computer at a frequency of 2 Hz.

ICP was monitored using an intraparenchymal probe (Codman ICP MicroSensor; Codman & Shurtleff, Inc., Raynham, MA). ABP was measured from the radial or femoral artery with an intravascular line connected to a pressure transducer (Baxter Healthcare Corp. CardioVascular Group, Irvine, CA).

Signals were digitized using an A/D converter (DT9801; Data Translation, Marlboro, MA), sampled at a frequency of 50 Hz. All data acquisition was done using a bedside laptop PC running ICM+ software (Cambridge Enterprise, University of Cambridge, UK).13

Back to Top | Article Outline

Data Analysis

Data Selection

Analysis of slow waves is easily confounded by external artifacts. For instance, increases of ICP caused by tracheal suctioning or positioning of the patient will result in high power in the slow wave frequency range. In contrast, here we aimed to investigate spontaneous slow fluctuations in the ICP and THI signals, and therefore only episodes not containing external artifacts were selected manually from the continuous recordings of the 37 patients. Selection criteria were as follows: (1) artifact-free signal of ICP, THI, and ABP—all episodes containing movement artifacts, “artificial” ICP increases caused by positioning or tracheal suctioning, or steep and sudden changes in any of the signals were excluded; and (2) minimum duration of 20 minutes.

Back to Top | Article Outline

Calculation of PRx and THx

PRx and THx were calculated as moving correlation coefficients, using 300-second time windows, between 10-second averaged values of ICP and ABP, and THI (right and left sides) and ABP, respectively.4,6 Positive PRx and THx values indicate worse pressure reactivity, whereas zero or negative values indicate better pressure reactivity.

Back to Top | Article Outline

Spectral Analysis

Welch's method (taking the average of periodograms calculated using fast Fourier transform of 50% overlapped data segments pretreated with Hamming window) was used for spectral analysis of ICP, THI, and ABP signals. Before spectral analysis, data were detrended using a DC filter (transition bandwidth 0.0008 Hz) to eliminate ultraslow fluctuations and baseline drifts of the signals. Spectral power of ICP, THI (average of both sides), and ABP was calculated every 10 seconds for the slow wave frequency range (0.005–0.055 Hz), using a 360-second time window. To compensate for varying levels of signal-to-noise ratio of the measured signals (where “noise” is defined as any component contributing to power outside of the targeted slow wave range) and to enable direct comparisons between patients, normalized spectral power values were calculated by dividing the power in the frequency band between 0.015 and 0.055 by the total power with the range of 0.01 and 0.5 Hz.

Back to Top | Article Outline

Statistics

Histogram plots were used to show distributions of power variables for the pooled data of all patients. Additionally, histograms showing the distribution of normalized slow-power values of THI and ICP for each subject individually are provided as supplemental files (see Supplemental Digital Content 1, http://links.lww.com/AA/A294, and 2, http://links.lww.com/AA/A295). For between-subject comparisons, the data were averaged per patient. Pearson correlation was used to assess the agreement between PRx and THx as a function of normalized power of slow oscillations in the input signals. For within-subject comparisons, the data analyzed for each patient individually and correlations between the different autoregulation indices were calculated as follows: First, to assess agreement between measurements over the right and left frontal region, THx values obtained from the right and left sides were compared for each patient individually. Second, the correlation between PRx and THx (based on the averaged THI signal from both sensors) was assessed for each patient individually during (a) episodes with high versus (b) low slow wave power of THI. The correlation coefficients were then compared between conditions (low versus high slow wave power of THI) using the Friedman test (please compare last paragraph of the Results section). Correlation coefficients and 95% confidence intervals were calculated using Matlab (version R2007b; MathWorks, Natick, MA). The Friedman test was performed using SAS software (SAS Institute, Cary, NC).

Back to Top | Article Outline

RESULTS

Patient Characteristics

The median age of the 37 patients was 34 years (interquartile range [IQR] 28.5, range 16–78). The median Glasgow coma scale score on admission was 7 (IQR 4.25, range 2–15). Nine patients (24.3%) had a favorable outcome (Glasgow outcome scale score of 4 or 5 at 3 months), and 7 patients died (18.9%). For further details and CT data, please refer to Table 1.

Back to Top | Article Outline

Episodes of High Versus Low Power of Slow Oscillations

Figure 1 shows an example of 2 artifact-free episodes derived from 1 patient. During the first episode, slow oscillations were clearly visible in the ICP and THI signals (Fig. 1A). Coherence between ICP and THI was highest in the slow frequency range at approximately 0.035 Hz; a second peak at 0.3 Hz represents the respiratory component. In contrast, during the second episode (Fig. 1B), the power of spontaneous low frequency oscillations was low. Only a single peak of coherence between ICP and THI was apparent within the respiratory frequency range (approximately 0.3 Hz). The presence of a respiratory peak might serve as an indicator for data quality. In case of low data quality because of technical difficulties, we would not expect a respiratory peak or high coherence of THI and ICP in the respiratory frequency range, because only noise, a random signal, would be sampled. However, according to our hypothesis, during episodes as shown in Figure 1B, assessment of cerebrovascular reactivity may not be applicable because of the absence of slow frequency oscillations.

Figure 1

Figure 1

Back to Top | Article Outline

Spectral Analysis of ICP, THI, and ABP

Figure 2 shows the distribution of normalized power of slow frequency oscillations of ICP, THI, and ABP based on the nonaveraged and pooled data of all patients. A bimodal distribution with phases of high versus low slow wave activity can clearly be distinguished for normalized power of ICP (median 0.56, IQR 0.35) and THI (median 0.42, IQR 0.37), but not for ABP (median 0.28, IQR 0.26). The individual distributions for each patient are provided as supplemental files (see Supplemental Digital Content 1, http://links.lww.com/AA/A294, and 2, http://links.lww.com/AA/A295).

Figure 2

Figure 2

Back to Top | Article Outline

Correlation Analysis Between PRx and THx as a Function of Slow Wave Power

Figure 3 shows the agreement between PRx and THx as a function of normalized power of slow oscillations in the input signals (THI and ICP). Each point represents the averaged data of 1 patient. The correlation between PRx and THx increased with increasing power of slow waves in the input signals. A correlation of r = 0.99 (95% confidence interval [CI] 0.51–1.00, P = 0.01) was found when normalized THI slow wave power was >0.6, as compared with r = 0.39 (95% CI −0.37 to 0.84, P = 0.30) for slow wave power <0.2. The respective correlations were r = 0.95 (95% CI 0.79–0.99, P < 0.01) versus r = 0.56 (95% CI −0.64 to 0.97) for normalized slow power of ICP. However, because of the small number in each category, it may be preferable to select only 2 categories; for patients with average normalized slow power >0.4 (categories 3 and 4), the agreement between THx and PRx was r = 0.80 (95% CI 0.53–0.92, P < 0.01). This threshold would allow inclusion of 52.9% of the data. The correlation in the remaining files was r = 0.07 (95% CI −0.40 to 0.51, P = 0.79).

Figure 3

Figure 3

Back to Top | Article Outline

Within-Subject Comparisons

Table 1 shows the correlations between indices as well as clinical and radiological data for each patient separately. Patients were ordered based on level of agreement between the THx values calculated from the data collected over the right and left frontal region (THx right versus THx left). This was done because of the consideration that PRx is a global parameter based on ICP whereas THx in contrast is derived from a local measurement, thereby possibly providing information on the local state of autoregulation. It therefore could be argued that, when there is good agreement of THx on both sides, a comparison of PRx and THx can be made. The level of agreement between the THx indices of both sides ranged from a high correlation (r = 0.75, P < 0.01) to no correlation (r = −0.04). The median correlation in our dataset was at 0.36 (dotted line).

Based on the threshold found for between-patient comparisons, for each patient, the correlation between PRx and THx was calculated during episodes of high and low normalized slow power of THI (nTHI >0.4 vs nTHI ≤0.4). Eight patients of our sample did not have episodes with normalized slow power of THI exceeding the threshold of 0.4. In contrast, for 1 patient, there were only data available for the high-power criterion.

The individual PRx-THx correlations during high and low slow-power activity were compared separately for the group of patients with a right-left correlation above the median (n = 19) versus below the median (n = 18). In the group of higher agreement between the THx values of both sides, the mean correlation between PRx and THx was 0.66 (IQR 0.30) for nTHI >0.4 compared with 0.35 (IQR 0.35) for nTHI ≤0.4 (Friedman, P = 0.004). This finding is in line with the hypothesis generated from the comparison between subjects. In contrast, in the group of patients with lower or no agreement between the THx values of both sides, there was a nonsignificant inverse relationship: the mean correlation between PRx and THx for episodes with nTHI >0.4 was 0.26 (IQR 0.67) compared with 0.34 (IQR 0.25) for episodes with nTHI ≤0.4.

Back to Top | Article Outline

DISCUSSION

Our results suggest that the agreement between the PRx and the newly introduced index THx is a function of the power of slow oscillations in the input signals. This finding confirms the intuitive notion that adequate assessment of cerebrovascular reactivity in general depends on the occurrence of low frequency oscillations in the respective input signals. Our findings thus corroborate the theoretical assumptions underlying noninvasive assessment of cerebrovascular reactivity. Furthermore, our data suggest that there should be a certain level of agreement between the THx measures on both sides to assume that THx provides information on the global state of autoregulation.

PRx has been validated in several clinical studies and a clear link to clinical outcome has been established.6,7,1416 However, continuous monitoring of autoregulation based on spontaneous slow oscillations has certain limitations. In contrast to autoregulation testing methods such as pharmacological increases in ABP, thigh cuff release, or tilt table declination test, continuous monitoring relies on the observation of spontaneous responses of cerebral blood flow (CBF) to spontaneous fluctuations in ABP, leading to a decreased signal-to-noise ratio. Furthermore, these variables are vulnerable to confounding influences such as CO2 changes or application of drugs, decreasing precision of the method. Averaging repeated measures over time aims to reduce this bias to provide a continuous estimation of autoregulation without the need of potentially harmful hemodynamic stimuli.17

THx has only been investigated in 1 clinical study performed by our group and including the same pool of patients. In the previous study, we reported a correlation of 0.56 between PRx and THx4 without, however, considering the presence of slow oscillations or side-to-side differences. Furthermore, there is still uncertainty about the applicability of NIRS in patients with brain injury.18 Data quality in patients with frontal contusions, brain swelling, or intracranial or subdural hemorrhages has been questioned.18,19 In addition, subdural air after craniotomy may render NIRS readings less reliable.20 Interestingly, as can be seen in Table 1, analysis of agreement between the indices derived from THI data from the right and left hemispheres did not depend on underlying pathology. Four of the 18 patients with high to moderate agreement between indices of both sides had frontal contusions or hemorrhages. In the aforementioned previous study, it was suggested that patients with frontal contusions be excluded, which improved the correlation between PRx and THx.4 However, the current analysis suggests that THx may be used as a substitute for PRx in patients with frontal contusions or hemorrhages, under the conditions that there is an at least moderate correlation in measurement of both sides and slow wave power in the input signal is sufficiently high. Instead of a priori exclusion of these patients, comparison of measurements of both sides and estimation of slow wave power could help in deciding whether THx may be used to help guide treatment in these patients.

Slow oscillations of cerebral hemodynamics are a well-known phenomenon and have been studied using different techniques. However, their origin and pathophysiological basis are still controversial. Using ICP recordings, Lundberg21 was the first to describe slow, non–heart beat– related oscillations with a frequency of 0.5 to 2.0 per minute, terming them B waves. Although he attributed the generation of B waves to concomitant CO2 fluctuations, other mechanisms for their generation have been proposed. Some groups postulated the existence of an autonomic brainstem pacemaker triggering ICP and simultaneous cardiovascular oscillations.22,23 Magnaes24 finally was among the first to suggest that B waves, more generally termed slow oscillations, might be generated by changes in intracranial blood volume, reflecting both blood pressure waves and cerebral autoregulation. The hypothesis of slow oscillations mirroring changes in the cerebral vascular tone in response to changes in CPP, and thereby ultimately reflecting changes in CBV, provides the pathophysiological basis for assessment of cerebrovascular pressure reactivity. Apart from ICP, rhythmic slow oscillations have been observed in other modalities, such as in cerebral flow velocities assessed by transcranial Doppler,25 cerebral oxygenation, and ABP.26 In addition, a number of studies have reported the occurrence of slow oscillations in NIRS signals in TBI,27 ischemic stroke patients,28 and the visual cortex of healthy controls,29,30 repeatedly linking them to vasomotion.27,29,30 Furthermore, persuasive evidence for the link between THI and CBV has been provided in an earlier experimental study in dogs.8 Assessing the effects of graded hypotension on CBF and CBV using microsphere-determined CBF and clearance of indocyanine green to assess mean transit time, the amount of total hemoglobin assessed by NIRS has been shown to reflect changes in CBV (CBF × mean transit time). Our findings and the experimental data of Lee et al.,5 showing a high coherence between ICP and THI in the slow wave spectrum, provide further, albeit indirect, evidence for this relationship, at the same time cross-validating the long-established index PRx by strengthening the link between ICP and THI, and ultimately CBV.

Limitations of our study include the retrospective design and the availability of variable amounts of data per patient, with 8 patients not meeting the slow wave criteria, possibly introducing bias. Finally, the validation against PRx only provides indirect evidence for the clinical value of THx. The latter, however, was not the intent of the current study and has been partially described before.4 Based on the same dataset, Zweifel et al.4 have shown that THx may be used to calculate optimal ABP, representing the noninvasive equivalent for optimal CPP and thereby providing a tool for blood pressure management in brain-injured patients. In the current study, we simply aimed to identify situations in which THx is most likely to be a noninvasive equivalent of PRx. As outlined above, PRx has certain shortcomings but has been validated in different clinical studies and shown to have implications for patient outcome and guidance of CPP, which led to inclusion of the index into current guidelines.31 A noninvasive equivalent for PRx therefore might be of clinical value for patients in whom invasive ICP monitoring is not applicable, such as children, patients with mild to moderate TBI, or for intraoperative monitoring during cardiac surgery to reduce risks of subsequent neurological deficits.32 Although our results suggest that approximately 50% of the monitoring data would have to be rejected because of absence of sufficient slow wave power, THx could provide a helpful tool in these patients. Furthermore, side-to-side comparisons provide an easy tool to estimate whether THx measurements provide more than local information. As a next step, considering the findings of the current study, NIRS-based assessment of autoregulation and ABP management should be validated in larger patient populations, and a direct link to outcome, as described for PRx, should be established.

Back to Top | Article Outline

CONCLUSION

We found that the correspondence between the validated measure of cerebrovascular reactivity PRx and the NIRS-based measure THx was a function of power of slow oscillations in the input signals. Furthermore, the agreement between THx measurements over both hemispheres can be used to estimate applicability of THx as a global measure. While NIRS-based cerebrovascular reactivity monitoring awaits further assessment in larger cohorts, our findings may be used as selection criteria to estimate reliability of the acquired data.

Back to Top | Article Outline

AUTHORS' CURRENT AFFILIATIONS

Jennifer Diedler, MD, is currently affiliated with the Department of Neurology, University of Heidelberg, Heidelberg, Germany; Christian Zweifel, MD, is currently with the University of Basel; Magdalena Kasprowicz, PhD, is currently with the Institute of Biomedical Engineering and Instrumentation, Wroclaw University of Technology, Poland; Enrico Sorrentino, MD, is currently with Wroclaw University; Christina Haubrich, MD, is currently with the University of Aachen; and Kenneth M. Brady, MD, is currently with the University of Texas.

Back to Top | Article Outline

FUNDING

JD was supported by a scholarship from the Medical Faculty, University of Heidelberg. This work was supported by the National Institute of Health Research, Biomedical Research Centre (Neuroscience Theme), the Medical Research Council (grants G0600986 and G9439390), NIHR Senior Investigator Awards (JDP), the Swiss National Science Foundation (PBBSP3-125550 to CZ, Bern, Switzerland, and the Foundation for Polish Science (MK). CZ received a travel grant from Hamamatsu Photonics, Welwyn Garden City, Hertfordshire, UK.

Back to Top | Article Outline

DISCLOSURES

Name: Jennifer Diedler, MD.

Contribution: This author helped design the study, conduct the study, and write the manuscript.

Attestation: Jennifer Diedler has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.

Conflicts of Interest: Jennifer Diedler reported no conflicts of interest.

Name: Christian Zweifel, MD.

Contribution: This author helped conduct the study.

Attestation: Christian Zweifel has seen the original study data and approved the final manuscript.

Conflicts of Interest: Christian Zweifel reported no conflicts of interest.

Name: Karol P. Budohoski, MD.

Contribution: This author helped design the study and analyze the data.

Attestation: Karol P. Budohoski has seen the original study data and approved the final manuscript.

Conflicts of Interest: Karol P. Budohoski reported no conflicts of interest.

Name: Magdalena Kasprowicz, PhD.

Contribution: This author helped design the study and analyze the data.

Attestation: Magdalena Kasprowicz has seen the original study data and approved the final manuscript.

Conflicts of Interest: Magdalena Kasprowicz reported no conflicts of interest.

Name: Enrico Sorrentino, MD.

Contribution: This author helped analyze the data.

Attestation: Enrico Sorrentino has seen the original study data and approved the final manuscript.

Conflicts of Interest: Enrico Sorrentino reported no conflicts of interest.

Name: Christina Haubrich, MD.

Contribution: This author helped design the study.

Attestation: Christina Haubrich has seen the original study data and approved the final manuscript.

Conflicts of Interest: Christina Haubrich reported no conflicts of interest.

Name: Kenneth M. Brady, MD.

Contribution: This author helped write the manuscript.

Attestation: Kenneth M. Brady approved the final manuscript.

Conflicts of Interest: Kenneth M. Brady reported no conflicts of interest.

Name: Marek Czosnyka, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Marek Czosnyka has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: The software for brain monitoring ICM+ (www.neurosurg.cam.ac.uk/icmplus) is licensed by Cambridge Enterprise Limited (University of Cambridge). Peter Smielewski and Marek Czosnyka have financial interest in a fraction of the licensing fee.

Name: John D. Pickard, FMedSci.

Contribution: This author helped with critical revision of the manuscript.

Attestation: John D. Pickard approved the final manuscript.

Conflicts of Interest: John D. Pickard reported no conflicts of interest.

Name: Peter Smielewski, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Peter Smielewski has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.

Conflicts of Interest: The software for brain monitoring ICM+ (www.neurosurg.cam.ac.uk/icmplus) is licensed by Cambridge Enterprise Limited (University of Cambridge). Peter Smielewski and Marek Czosnyka have financial interest in a fraction of the licensing fee.

This manuscript was handled by: Gregory J. Crosby, MD.

Back to Top | Article Outline

ACKNOWLEDGMENTS

Many thanks to all staff of the Neurocritical Care Unit, Addenbrooke's Hospital, Cambridge, UK, for help and support during data collection.

Back to Top | Article Outline

REFERENCES

1. Steiner LA, Pfister D, Strebel SP, Radolovich D, Smielewski P, Czosnyka M. Near-infrared spectroscopy can monitor dynamic cerebral autoregulation in adults. Neurocrit Care 2009;10:122–8
2. Wagner BP, Pfenninger J. Dynamic cerebral autoregulatory response to blood pressure rise measured by near-infrared spectroscopy and intracranial pressure. Crit Care Med 2002;30:2014–21
3. Nissen P, Pacino H, Frederiksen HJ, Novovic S, Secher NH. Near-infrared spectroscopy for evaluation of cerebral autoregulation during orthotopic liver transplantation. Neurocrit Care 2009;11:235–41
4. Zweifel C, Castellani G, Czosnyka M, Helmy A, Manktelow A, Carrera E, Brady KM, Hutchinson PJ, Menon DK, Pickard JD, Smielewski P. Noninvasive monitoring of cerebrovascular reactivity with near infrared spectroscopy in head-injured patients. J Neurotrauma 2010;27:1951–8
5. Lee JK, Kibler KK, Benni PB, Easley RB, Czosnyka M, Smielewski P, Koehler RC, Shaffner DH, Brady KM. Cerebrovascular reactivity measured by near-infrared spectroscopy. Stroke 2009;40:1820–6
6. Czosnyka M, Smielewski P, Kirkpatrick P, Laing RJ, Menon D, Pickard JD. Continuous assessment of the cerebral vasomotor reactivity in head injury. Neurosurgery 1997;41:11–7
7. Steiner LA, Coles JP, Johnston AJ, Chatfield DA, Smielewski P, Fryer TD, Aigbirhio FI, Clark JC, Pickard JD, Menon DK, Czosnyka M. Assessment of cerebrovascular autoregulation in head-injured patients: a validation study. Stroke 2003;34:2404–9
8. Ferrari M, Wilson DA, Hanley DF, Traystman RJ. Effects of graded hypotension on cerebral blood flow, blood volume, and mean transit time in dogs. Am J Physiol 1992;262:H1908–14
9. Patel HC, Menon DK, Tebbs S, Hawker R, Hutchinson PJ, Kirkpatrick PJ. Specialist neurocritical care and outcome from head injury. Intensive Care Med 2002;28:547–53
10. Marshall LF, Marshall SB, Klauber MR, Van Berkum Clark M, Eisenberg H, Jane JA, Luerssen TG, Marmarou A, Foulkes MA. The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 1992;9:S287–92
11. Al-Rawi PG, Smielewski P, Kirkpatrick PJ. Evaluation of a near-infrared spectrometer (NIRO 300) for the detection of intracranial oxygenation changes in the adult head. Stroke 2001;32:2492–5
12. Yoshitani K, Kawaguchi M, Miura N, Okuno T, Kanoda T, Ohnishi Y, Kuro M. Effects of hemoglobin concentration, skull thickness, and the area of the cerebrospinal fluid layer on near-infrared spectroscopy measurements. Anesthesiology 2007;106:458–62
13. Smielewski P, Czosnyka M, Steiner L, Belestri M, Piechnik S, Pickard JD. ICM+: software for on-line analysis of bedside monitoring data after severe head trauma. Acta Neurochir Suppl 2005;95:43–9
14. Zweifel C, Lavinio A, Steiner LA, Radolovich D, Smielewski P, Timofeev I, Hiler M, Balestreri M, Kirkpatrick PJ, Pickard JD, Hutchinson PJA, Czosnyka M. Continuous monitoring of cerebrovascular pressure reactivity in patients with head injury. Neurosurg Focus 2008;25:E2
15. Steiner LA, Czosnyka M, Piechnik SK, Smielewski P, Chatfield D, Menon DK, Pickard JD. Continuous monitoring of cerebrovascular pressure reactivity allows determination of optimal cerebral perfusion pressure in patients with traumatic brain injury. Crit Care Med 2002;30:733–8
16. Steiner LA, Coles JP, Czosnyka M, Minhas PS, Fryer TD, Aigbirhio FI, Clark JC, Smielewski P, Chatfield DA, Donovan T, Pickard JD, Menon DK. Cerebrovascular pressure reactivity is related to global cerebral oxygen metabolism after head injury. J Neurol Neurosurg Psychiatry 2003;74:765–70
17. Czosnyka M, Brady K, Reinhard M, Smielewski P, Steiner LA. Monitoring of cerebrovascular autoregulation: facts, myths, and missing links. Neurocrit Care 2009;10:373–86
18. Calderon-Arnulphi M, Alaraj A, Slavin KV. Near infrared technology in neuroscience: past, present and future. Neurol Res 2009;31:605–14
19. Robertson CS, Gopinath SP, Chance B. A new application for near-infrared spectroscopy: detection of delayed intracranial hematomas after head injury. J Neurotrauma 1995;12:591–600
20. Büchner K, Meixensberger J, Dings J, Roosen K. Near-infrared spectroscopy: not useful to monitor cerebral oxygenation after severe brain injury. Zentralbl Neurochir 2000;61:69–73
21. Lundberg N. Continuous recording and control of ventricular fluid pressure in neurosurgical practice. Acta Psychiatr Scand Suppl 1960;36:1–193
22. Einhaupl K, Garner C, Dirnagl U, Schmieder G, Schmiedek P, Kufner G, Rieder J. Oscillations of ICP related to cardiovascular parameters. Berlin: Springer-Verlag, 1986
23. Venes JL. B-waves: a reflection of cardiorespiratory or cerebral nervous systems rhythm? Childs Brain 1979;5:352–60
24. Magnaes B. Body position and cerebrospinal fluid pressure. Part 2. Clinical studies on orthostatic pressure and the hydrostatic indifferent point. J Neurosurg 1976;44:698–705
25. Newell DW, Aaslid R, Stooss R, Reulen HJ. The relationship of blood flow velocity fluctuations to intracranial pressure B waves. J Neurosurg 1992;76:415–21
26. Steinmeier R, Bauhuf C, Hubner U, Bauer RD, Fahlbusch R, Laumer R, Bondar I. Slow rhythmic oscillations of blood pressure, intracranial pressure, microcirculation, and cerebral oxygenation: dynamic interrelation and time course in humans. Stroke 1996;27:2236–43
27. Cheng OS, Prowse S, Strong AJ. Oscillations in the near-infrared signal in patients with severe head injury. Acta Neurochir Suppl 2002;81:135–7
28. Li Z, Wang Y, Li Y, Li J, Zhang L. Wavelet analysis of cerebral oxygenation signal measured by near infrared spectroscopy in subjects with cerebral infarction. Microvasc Res 2010;80:142–7
29. Obrig H, Neufang M, Wenzel R, Kohl M, Steinbrink J, Einhaupl K, Villringer A. Spontaneous low frequency oscillations of cerebral hemodynamics and metabolism in human adults. Neuroimage 2000;12:623–39
30. Elwell CE, Springett R, Hillman E, Delpy DT. Oscillations in cerebral haemodynamics: implications for functional activation studies. Adv Exp Med Biol 1999;471:57–65
31. Brain Trauma Foundation; American Association of Neurological Surgeons; Congress of Neurological Surgeons; Joint Section on Neurotrauma and Critical Care AANS/CNS, Bratton SL, Chestnut RM, Ghajar J, McConnell Hammond FF, Harris OA, Hartl R, Manley GT, Nemecek A, Newell DW, Rosenthal G, Schouten J, Shutter L, Timmons SD, Ullman JS, Videtta W, Wilberger JE, Wright DW. Guidelines for the management of severe traumatic brain injury. IX. Cerebral perfusion thresholds. J Neurotrauma 2007;24:S59–64
32. Brady K, Joshi B, Zweifel C, Smielewski P, Czosnyka M, Easley RB, Hogue CWJ. Real-time continuous monitoring of cerebral blood flow autoregulation using near-infrared spectroscopy in patients undergoing cardiopulmonary bypass. Stroke 2010;41:1951–6

Supplemental Digital Content

Back to Top | Article Outline
© 2011 International Anesthesia Research Society