SEARCH

SEARCH BY CITATION

Keywords:

  • Brain: oxygenation;
  • Measurement techniques: infra-red;
  • Monitoring: oxygen.

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

In this study cerebral oxygenation was measured using the NIRO 300 and the INVOS 5100 spectrophotometers in 10 healthy adult volunteers, exposed to varying degrees of hyperoxia and hypoxia. The results showed similar baseline values for tissue oxygenation index and regional cerebral oxygen saturation with mean (SD) values being 64.9% (5.1) and 62.3% (6.0), respectively. The overall bias was –2.1%, with the INVOS 5100 under-reading cerebral oxygenation compared to the NIRO 300, with limits of agreement of ±14.7%. Both monitors demonstrated similar changes in response to hyperoxia and hypocapnia (coefficient of variance for FIo2 0.45 = 10.0%, FIo2 1.0 = 10.1%, hypocapnia = 14.5%). The reasons for the bias and variability may relate to differences in the methodological approaches of the two monitors. The correlation between the monitors in response to changes in cerebral oxygenation implies that they may be useful as trend monitors in clinical practice.

An easy to use, reliable and non-invasive monitor of cerebral oxygenation would be a useful tool to guide clinical interventions on the neuro-intensive care unit. Near-infra-red spectroscopy (NIRS) has previously been used to monitor changes in cerebral oxygenation and haemodynamics in neonates and adults [1, 2]. NIRS systems to date have predominantly displayed parameters, such as changes in oxygenated and deoxygenated haemoglobin concentrations, which are unfamiliar to clinicians. However, recent developments in NIRS technology have seen the advent of instruments that display a measure of absolute cerebral saturation in an attempt to improve user friendliness and aid clinical decision making. The NIRO 300 (Hamamatsu Photonics, Hamamatsu City, Japan) and the INVOS 5100 (Somanetics, Troy, MI, USA) both measure cerebral tissue oxygenation and return a single numerator, tissue oxygenation index (TOI) from the NIRO 300 and regional cerebral oxygen saturation (rso2) from the INVOS 5100. These instruments differ significantly in the methodologies and assumptions used to calculate their cerebral oxygenation indices [3–5] and therefore the interchangeability of the two monitors is not clear.

For the clinician attempting to measure cerebral oxygenation at the bedside, interpretation of a single variable is preferred. However, the clinical application of these new devices will clearly depend on their validity, reliability and reproducibility. The aim of this study was to determine whether the NIRO 300 and INVOS 5100 were able to detect changes in cerebral oxygenation during hyperoxia and hyperventilation in healthy volunteers, and to compare the values recorded from each system. We also assessed qualitatively how easy and reliable these monitors are to use.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

Following local ethics committee approval and written informed consent, 10 healthy adult volunteers were studied.

Description of spectrophotometers

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

The physical principles upon which NIRS is based have been well documented previously [6]. In general terms, the change in concentration of coloured compounds (chromophores) within the field of view is measured using modifications of the Beer–Lambert law. Oxygenated and deoxygenated haemoglobin absorb near infra-red light at specific wavelengths [7] and the degree of change in light absorption is related to the change in concentration of that chromophore within the tissue illuminated. However, to determine the absolute concentration of chromophore present, the amount of light lost to scatter needs to be known [7].

The NIRO 300 uses laser-emitting diodes to generate light at four different wavelengths, typically 775, 810, 850 and 910 nm. It employs the technique of spatially resolved spectroscopy (SRS) and uses multiple, closely spaced detectors to measure light attenuation as a function of source–detector separation. From these measurements, it is possible to de-couple the absorption and scattering coefficients and combine this with an estimate of the wavelength dependence of light scattering. The result is a measurement of scaled absolute haemoglobin concentrations, i.e. the relative proportions of oxyhaemoglobin and deoxyhaemoglobin, from which tissue oxygen saturation is easily computed [8]. No adjustment is made for extracerebral blood and no assumption is made regarding the arterial-to-venous partition ratio. The interoptode distance is user adjustable but the manufactured optode holder gives a choice of 4 cm or 5 cm separations. A 5-cm interoptode separation was chosen for this study as this provided an optimal signal-to-noise ratio. The sampling frequency is also variable, from once per minute up to 6 Hz, with measurements averaged over the selected time period. A sampling time of 10 s was chosen for this study. In addition to measuring TOI, the NIRO 300 also continuously measures, displays and stores changes in oxyhaemoglobin and deoxyhaemoglobin, oxidised cytochrome oxidase and total haemoglobin (THI) concentrations.

The INVOS Cerebral Oximeter generates two wavelengths of light (730 and 810 nm) from light-emitting diodes (LEDs) which are alternately illuminated. The INVOS 5100 utilises a proprietary ‘disposable’ sensor, incorporating the LEDs and two light-collecting optodes at fixed distances, for adult use of 3 cm and 4 cm, from the emitter. The INVOS system uses a simplified form of SRS, which includes some subtle differences to the approach used in the NIRO 300. Light attenuation measurements are made as a function of spacing across two detectors rather than the three used in the NIRO 300. In the INVOS, it is assumed that there is no wavelength dependence of scattering and that there is linearity over the 1 cm between the two detectors [9]. The sensor is designed to be placed only on the frontal region. In common with other organs, the brain is assumed to have a greater volume of venous blood than arterial blood. The relative contribution of arterial and venous volumes to NIRS-derived signals cannot be measured in real-time and may change in response to other physiological changes such as an alteration in Paco2. The INVOS 5100 assumes an unchanging partition ratio of 25%:75% for arterial:venous blood, based on studies of the cerebral circulation [10, 11], and uses this ratio when calculating its values for rso2. It is unable to quantify light scatter, but makes the assumption that the degree of scatter is constant (although unknown) throughout the measurement period and can effectively be ignored. Light absorption data is collected continuously 15 times a second. When 50 samples have been collected (3.3 s) they are averaged to determine a new rso2 value and the display on the monitor is updated. The INVOS 5100 displays only a value for rso2.

Both monitors have the facility for simultaneous dual channel recording.

Study design

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

The optodes of the NIRO 300 and INVOS 5100 were randomly placed on opposite sides of the forehead, 1 cm above the eyebrow, 1 cm lateral to the midline and distant from the temporalis muscle. The sensors were held in place with a loose crepe bandage and covered by a light shielding cloth.

Arterial oxygen saturation (Spo2), end-tidal carbon dioxide tension (FEco2), rso2 and TOI were measured continuously throughout the study and the data recorded manually every 15 s. The study had four sequential stages shown schematically in Fig. 1. The volunteers were first asked to breathe normally in room air (baseline) and then to breathe hyperoxic gas mixtures of FIo2 = 0.45 followed by FIo2 = 1.0. The FEco2 was maintained at normal levels using an anaesthetic circle system. The volunteers were finally asked to hyperventilate to FEco2 of 3.0 kPa in room air. The first 3 min of each stage was used as a period for equilibration at the new FIo2 or FEco2. This was followed by a 5-min data collection period before moving to the next stage of the study. The ease of use of the two monitors was also noted.

image

Figure 1. Schematic representation of the study protocol.

Download figure to PowerPoint

Statistical analysis

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

The data were analysed using SPSS 10.1. The mean value of the serial measurements of the recorded variables during the 5-min data collection period was used as the summary value [12]. Changes in Spo2 and FEco2 were assessed using a paired t-test. The Mann–Whitney U-test for unpaired samples was used to compare the output from each monitor, and the intrasubject changes in rso2 and TOI with hyperoxia and hyperventilation were compared using Wilcoxon matched pairs tests and their coefficient of variance calculated.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

Eight male and two female volunteers were studied with a mean (range) age of 33.9 (28–37) years. There was no change in Spo2 from baseline to FIo2 = 0.45. With increase of FIo2 from baseline to 1.0, mean Spo2 rose from 98.0 to 98.9% (p = 0.04). During hyperventilation, mean Spo2 increased from 98.0 to 98.8% (p = 0.04). There was no significant change in mean FEco2 with increasing FIo2 but there was a significant reduction from the baseline value of 4.3 kPa to 2.9 kPa (p < 0.001) during hyperventilation (Table 1).

Table 1.  Mean and standard deviation (SD) values for Spo2 and FEco2 during spontaneous ventilation in room air (baseline), 0.45 FIo2, 1.0 FIo2 and during hyperventilation. n = 10.
InterventionSpo2(%)SignificanceFEco2(kPa)Significance
Baseline98.0 (0.75)4.3 (0.6) 
45% O298.6 (0.88) p = 0.094.3 (0.5)NS
100% O298.9 (0.94) p = 0.044.3 (0.5)NS
Hyperventilation98.8 (0.77) p = 0.042.9 (0.1)p < 0.001

The baseline values for TOI and rso2 are shown in Fig. 2. The mean (SD) values were 64.9% (5.1) and 62.3% (6.0) for TOI and rso2, respectively (p = 1.0), with substantial intersubject variation in the baseline. The coefficient of variance for absolute baseline values was 9.38%.

image

Figure 2. n  = 10. Comparison of absolute values for: (a) rso2 (INVOS 5100) and (b) TOI (NIRO 300) labelled as percentage saturation, displayed for each volunteer when breathing room air, with a mean FEco2 of 4.3 kPa (baseline). No significant difference was noted between the mean values obtained from the two monitors. Mean and standard deviation (SD) for the INVOS was 62.3% (6.0) vs. NIRO 64.9% (5.1), p = 1.00 Mann–Whitney U-test.

Download figure to PowerPoint

To examine the agreement between the two monitors, a Bland–Altman relationship [13] was plotted (Fig. 3). rso2 and TOI data collected during each stage of the study were included in this analysis. Summary values of rso2 are 2.1% lower than TOI and the standard deviation of the difference between the two monitors is 7.5%. The ±1.96 standard deviation limit of agreement between the two monitors is 14.7%.

image

Figure 3. n  = 10. Bland Altman plot comparing the INVOS 5100 and NIRO 300. The difference between cerebral oxygenation indices for the INVOS and NIRO are plotted against their mean saturation value for each subject under conditions of normocapnia breathing room air (baseline [×], 45% Oxygen [▵] 100% Oxygen [+]) and hypocapnia breathing room air [□]). Bias of −2.06% demonstrates that the INVOS under-reads cerebral oxygenation when compared to the NIRO. Note the large limits of agreement (1.96*SD = 14.66) indicative of the high degree of intra- and interindividual variability between the two indices, independent of changes in FIo2 or FEco2.

Download figure to PowerPoint

We also compared the difference from baseline for each individual during hyperoxia and hyperventilation (Fig. 4). There was no change from baseline TOI at FIo2 = 0.45 and a significant increase of 3.6% (p = 0.022) at FIo2 1.0. There was a significant increase in rso2 of 2.3% and 4.5% at FIo2 of 0.45 and 1.0, respectively (p = 0.047 and 0.017). TOI fell by 5.5% (p = 0.007) and rso2 by 4.2% (p = 0.017) during hyperventilation. The overall variability associated with the NIRO 300 results was lower than that associated with the INVOS 5100.

image

Figure 4. n  = 10. Cerebral oxygenation represented as a change from each individual's baseline value in response to the three interventions (breathing 0.45 FIo2, 1.0 FIo2 and hyperventilation to a mean FEco2 of 2.9 kPa), for both monitors. The median, interquartile and ranges and outliers [□] for all values measured during each intervention are represented as Box plots. The INVOS measures a significant change from baseline in response to all three interventions (p < 0.05), whereas the NIRO demonstrates a significant change in response to 1.0 FIo2 (p = 0.022) and hyperventilation (p = 0.007).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

The values for rso2 and TOI are similar, but there are large inter- and intra-individual differences between the outputs from the two monitors. The overall bias of −2.1% with wide limits of agreement (1.96 SD limit of agreement of 14.7%) makes interpretation of the absolute values of TOI and rso2 difficult and adds to the debate over what is actually being measured [14]. Our results are similar to those of Yoshitani and colleagues, who recorded a bias of − 0.5% (1.96 SD limit of agreement of 15.6%) in a comparison of the INVOS 4100 and NIRO 300 during a CO2 challenge test in anaesthetised patients [15]. The wide limits of agreement suggest that the absolute values for rso2 and TOI may be too imprecise for clinical use. However, in our study, the intra-individual changes in rso2 and TOI in response to hyperoxia and hypocapnia were similar, suggesting that both the INVOS 5100 and NIRO 300 may have possible roles as trend monitors.

We have assumed that the hyperoxic challenge assessed whether the monitors were able to detect changes in cerebral oxygenation secondary to an increase in cerebral oxygen delivery. Both monitors detected a significant increase in cerebral oxygenation when the subjects were breathing 100% oxygen, although only the INVOS detected a significant increase with 45% oxygen. We have also assumed that the hypocapnic challenge led to a reduction in cerebral blood flow although we were not able to confirm this during our study. However, both monitors did detect a reduction in cerebral oxygenation during hypocapnia and this may reflect a reduction in cerebral blood flow and oxygen delivery. Both monitors measured a significant decrease in cerebral oxygenation during hypocapnia compared to baseline FEco2 values.

The small, but statistically significant, increase in mean Spo2 during FIo2 of 1.0 or during hyperventilation may have affected calculated cerebral oxygenation in this study. Cerebral oxygenation is also affected by many other variables. These include arterial, venous and capillary saturation, the partition ratio between the arterial and venous components of the cerebral vasculature, cerebral blood volume and possibly scalp blood flow [10, 16, 17]. It is possible that the bias and variability between the two monitors may in part be related to the different way in which these factors are included in the methodologies employed. The INVOS 5100 calculates rso2 using the modified Beer–Lambert law and, to improve accuracy, includes a compensation for extracranial blood. The system uses two different emitter–detector separations with the intention of subtracting out extracranial contribution to the displayed variable. One, at 3 cm, is assumed to interrogate mainly extracerebral tissue and the other, at 4 cm, is assumed to interrogate cortical and intracerebral structures. Subtraction of the first from the second is then assumed to represent only the intracranial compartment [18]. However, 4 cm may be an inadequate interoptode separation to exclude totally extracranial absorption [19, 20] and it may be that the subtraction algorithm employed is too simplistic [14]. The NIRO 300 relies on a refinement of the modified Beer–Lambert law and incorporates a more rigorous application of spatially resolved spectroscopy [5, 8]. Although it does not specifically attempt to exclude any contribution from extracranial blood, a degree of compensation does occur which is intrinsic to the algebraic differentiation of the attenuation gradient measured by the multichannel detector. Recently, using the NIRO 300 to investigate cerebral oxygenation changes during carotid surgery, the sensitivity of TOI to intracranial changes was 87.5% with a specificity of 100%[21]. In contrast, the sensitivity and specificity to extracranial changes were 0%, suggesting that TOI, as measured by the NIRO 300, reflects changes in cerebral tissue oxygenation.

The two monitors also employ different light emitters. The NIRO 300 uses laser diodes with a very narrow emission bandwidth corresponding to the absorption characteristics of oxygenated and deoxygenated haemoglobin. However, the INVOS 5100 uses light diodes, which have a much broader bandwidth (30–40 nm), and this may affect their accuracy [22].

Finally, the two monitors use different algorithms for the calculation of TOI and rso2. It has been shown that the same set of NIRS data can give rise to differences in calculated chromophore concentration when analysed using different published algorithms [23]. The TOI algorithm for the NIRO 300, based on photon diffusion theory, is available but has only been verified in animal and in vitro studies [8]. Although the basic subtraction principle used by the INVOS 5100 to compensate for extracranial absorption is well known, the full details of the algorithm used to determine regional cerebral oxygen saturation have not been published.

The two monitors differed subjectively in their ease of use. We found the INVOS 5100 fast and easy to set up, application of the adhesive optode patch was simple and did not require any further light shielding. However, the optode patch malfunctioned on two occasions out of 10, necessitating replacement of the whole patch. The NIRO 300 required its optodes to be initialised prior to use, manipulated into the preformed rubber holder and attached to the forehead using crepe bandaging, a process which is time consuming. After this was achieved, further adjustments to optimise light shielding and reduce emitted light intensity were often required before an acceptable signal was obtained. However, once set up, both machines proved reliable and easy to use.

The INVOS 5100 is relatively new to the market but the validity of its predecessor, the INVOS 3100, is still debated [24], with studies supporting and refuting its reliability for rso2 measurement when compared with jugular venous bulb oximetry [25] and other spectrophotometers [26, 27]. The NIRO 300 has also recently been released but its predecessor, the NIRO 500, has similarly been compared with jugular venous oximetry and measurements of brain tissue po2[28] and also with the INVOS 3100 [3, 27]. There are few published studies to validate TOI as an independent monitor of cerebral oxygenation [29] but the recent report demonstrating high sensitivity and specificity for intracranial changes is likely to stimulate interest in its use in the clinical setting [21]. However, validation of new devices is complicated by the absence of a ‘gold standard’ against which monitors of cerebral oxygenation can be tested at the bedside. Jugular venous oximetry is a global, rather than regional, monitor [30] and involves only the venous compartment. Tissue microprobes measure tissue po2 directly but within an extremely small volume which may not be representative of the whole brain [30]. Cerebral saturation, however, is a value derived from all vascular compartments and is a relatively focal measurement. It is possible that cerebral saturation might represent a more useful measure of cerebral oxygenation than other variables in the clinical setting. Another potential advantage of the NIRO 300 is its ability to measure also THI, a scaled estimate of cerebral blood volume. The combination of THI and TOI will provide a more complete picture of cerebral haemodynamics and oxygenation at the bedside.

This study demonstrates that the NIRO 300 and the INVOS 5100 are able to measure cerebral oxygenation but the large degree of intra- and intersubject variability makes their absolute values of uncertain clinical value. However, our results suggest that these monitors might usefully be applied in the clinical setting to measure trends in cerebral oxygenation. The non-invasive nature of these devices and the simplicity of their use are major advantages over other methods of measuring cerebral oxygenation. Further validation of the use of NIRS for non-invasive monitoring of cerebral oxygenation at the bedside is warranted.

Acknowledgments

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References

We are grateful to Hamamatsu Photonics for loan of the NIRO 300 and to Somanetics for loan of the INVOS 5100.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Description of spectrophotometers
  5. Study design
  6. Statistical analysis
  7. Results
  8. Discussion
  9. Acknowledgments
  10. References
  • 1
    Edwards AD, Wyatt JS, Richardson C, Delpy DT, Cope M, Reynolds EO. Cotside measurement of cerebral blood flow in ill newborn infants by near infrared spectroscopy. Lancet 1988; 2: 7701.
  • 2
    Elwell CE, Cope M, Edwards AD, Wyatt JS, Reynolds EO, Delpy DT. Measurement of cerebral blood flow in adult humans using near infrared spectroscopy – methodology and possible errors. Advances in Experimental Medicine and Biology 1992; 317: 235.
  • 3
    Gomersall CD, Leung PL, Gin T, et al. A comparison of the Hamamatsu NIRO 500 and the INVOS 3100 near-infrared spectrophotometers. Anaesthesia and Intensive Care 1998; 26: 54857.
  • 4
    McCormick PW, Stewart M, Goetting MG, Balakrishnan G. Regional cerebrovascular oxygen saturation measured by optical spectroscopy in humans. Stroke 1991; 22: 596602.
  • 5
    Matcher SJ, Kirkpatrick P, Nahid K, Cope M, Delpy DT. Absolute quantification methods in tissue near infrared spectroscopy. Proceedings SPIE 1995; 2389: 48695.
  • 6
    Owen-Reece H, Smith M, Elwell CE, Goldstone JC. Near infrared spectroscopy. British Journal of Anaesthesia 1999; 82: 41826.
  • 7
    Elwell CE. The physical principles of tissue spectroscopy. In: A Practical Users Guide to Near Infrared Spectroscopy. Japan: Hamamatsu Photonics KK, 1995.
  • 8
    Suzuki S, Takasaki S, Ozaki T, Kobayashi Y. A tissue oxygenation monitor using NIR spatially resolved spectroscopy. Proceedings SPIE 1999; 3597: 58292.
  • 9
    Kim M, Ward D, Cartwright C, Kolano J, Chlebowski S, Henson I. Estimation of jugular venous saturation from cerebral oximetry of arterial O2 saturation during isocapnic hypoxia. Journal of Clinical Monitoring 2001; 16: 1919.
  • 10
    Mchedlishvilli G. In: Arterial Behaviour and Blood Circulation in the Brain. New York: Plenum Press, 1986.
  • 11
    Henson LC, Cartwright C, Chlebowski SM. Estimation of jugular venous O2 saturation from cerebral oximetry or arterial O2 saturation during isocapnic hypoxia. Anesthesiology 1997; 87: A402.
  • 12
    Matthews JN, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. British Medical Journal 1990; 300: 2305.
  • 13
    Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 30710.
  • 14
    Harris DN, Bailey SM. Near infrared spectroscopy in adults. Does the Invos 3100 really measure intracerebral oxygenation? Anaesthesia 1993; 48: 6946.
  • 15
    Yoshitani K, Kawaguchi M, Tatsumi K, Kitaguchi K, Furuya H. A comparison of the INVOS 4100 and the NIRO 300 near-infrared spectrophotometers. Anesthesia and Analgesia 2002; 94: 58690.
  • 16
    Samra SK, Stanley JC, Zelenock GB, Dorje P. An assessment of contributions made by extracranial tissues during cerebral oximetry. Journal of Neurosurgical Anesthesiology 1999; 11: 15.
  • 17
    Watzman HM, Kurth CD, Montenegro LM, Rome J, Steven JM, Nicolson SC. Arterial and venous contributions to near-infrared cerebral oximetry. Anesthesiology 2000; 93: 94753.
  • 18
    McCormick PW, Stewart M, Goetting MG, Dujovny M, Lewis G, Ausman JI. Noninvasive cerebral optical spectroscopy for monitoring cerebral oxygen delivery and hemodynamics. Critical Care Medicine 1991; 19: 8997.
  • 19
    Duncan A, Meek JH, Clemence M, et al. Optical pathlength measurements on adult head, calf and forearm and the head of the newborn infant using phase resolved optical spectroscopy. Physics in Medicine and Biology 1995; 40: 295304.
  • 20
    Firbank M, Okada E, Delpy DT. A theoretical study of the signal contribution of regions of the adult head to near-infrared spectroscopy studies of visual evoked responses. Neuroimage 1998; 8: 6978.
  • 21
    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: 2492500.
  • 22
    McKeating EG, Monjardino JR, Signorini DF, Souter MJ, Andrews PJ. A comparison of the Invos 3100 and the Critikon 2020 near-infrared spectrophotometers as monitors of cerebral oxygenation. Anaesthesia 1997; 52: 13640.
  • 23
    Matcher SJ, Elwell CE, Cooper CE, Cope M, Delpy DT. Performance comparison of several published tissue near-infrared spectroscopy algorithms. Analytical Biochemistry 1995; 227: 5468.
  • 24
    Gomersall CD, Joynt GM, Gin T, Freebairn RC, Stewart IE. Failure of the INVOS 3100 cerebral oximeter to detect complete absence of cerebral blood flow. Critical Care Medicine 1997; 25: 12524.
  • 25
    Grubhofer G, Lassnigg A, Manlik F, Marx E, Trubel W, Hiesmayr M. The contribution of extracranial blood oxygenation on near-infrared spectroscopy during carotid thrombendarterectomy. Anaesthesia 1997; 52: 11620.
  • 26
    Colier WN, Van Haaren NJ, Oeseburg B. A comparative study of two near infrared spectrophotometers for the assessment of cerebral haemodynamics. Acta Anaesthesiologica Scandinavica Suppl 1995; 107: 1015.
  • 27
    Grubhofer G, Tonninger W, Keznickl P, Skyllouriotis P, Ehrlich M, Hiesmayr M, Lassnigg A. A comparison of the monitors INVOS 3100 and NIRO 500 in detecting changes in cerebral oxygenation. Acta Anaesthesiologica Scandinavica 1999; 43: 4705.
  • 28
    McLeod ADM, Smith M, Igielman F, Goldstone JC. Cerebral oxygenation during normocapnic hyperoxia: comparison of tissue microprobes, spatially resolved near infrared spectroscopy and jugular venous oximetry. British Journal of Anaesthesia 2000; 84: 2745.
  • 29
    Quaresima V, Sacco S, Totaro R, Ferrari M. Noninvasive measurement of cerebral hemoglobin oxygen saturation using two near infrared spectroscopy approaches. Journal of Biomedical Optics 2000; 5: 2015.
  • 30
    Gupta AK, Hutchinson PJ, Al-Rawi P, et al. Measuring brain tissue oxygenation compared with jugular venous oxygen saturation for monitoring cerebral oxygenation after traumatic brain injury. Anesthesia and Analgesia 1999; 88: 54953.