Cerebral hypoxic/ischemic injury results from a mismatch between cerebral oxygen supply and demand across a spectrum of disorders, and continuous real-time monitoring of the adequacy of cerebral oxygenation can provide important therapeutic information. Measurement of cerebral oxygenation is widely used to assess the balance between cerebral metabolic supply and demand, but standard methods have significant limitations.1 Time-critical windows to prevent or minimize permanent ischemic neurologic injury may therefore pass silently because detection of cerebral ischemia/hypoxia in real-time remains problematic.
Near-infrared spectroscopy (NIRS) is a noninvasive optical technique that has potential to address many of the shortcomings inherent to other cerebral monitoring modalities.2 However, for over 3 decades since its first description, its clinical adoption has been limited. There is a wealth of data describing the use of NIRS in neonates where it has found clinical application since the 1980s,3 not least because invasive brain monitoring is not possible in this patient group.4,5 NIRS has also been widely used in nonbrain tissue6 and during functional brain imaging,7 but clinical data from adult brain monitoring are limited.
This review will examine the principles of cerebral NIRS, its usefulness, and limitations in adult brain monitoring, and will highlight potential future directions.
PRINCIPLES OF NIRS
Although a comprehensive review of the physical principles underlying the clinical use of NIRS is beyond the scope of this article (for this, the reader is referred to comprehensive reviews on the subject),8,9 an understanding of these principles is essential for clinicians to interpret NIRS data and navigate the plethora of commercially available devices.
The NIRS technique was first described in 1977 by Franz Jöbsis who made 2 key observations regarding near-infrared (NIR) light. First, light in the NIR spectrum (700–950 nm) can traverse biological tissue because of the relative transparency of tissue to light in this wavelength range and, second, several biological molecules, termed chromophores, have distinct absorption spectra in the NIR.10 From a clinical perspective, oxyhemoglobin (O2Hb) and deoxyhemoglobin (HHb) are the most commonly measured chromophores, although cytochrome c oxidase (CCO) may prove clinically more important11,12 and was in fact the target of Jöbsis’s initial investigations.
NIRS is based on the transmission and absorption of NIR light as it passes through tissue, and there are basic features common to all NIRS devices.9 Light is generated at specific wavelengths typically by light-emitting diodes, and is usually detected by silicon photodiodes.13 Alternative methods of light detection include photomultiplier tubes in older devices, and charged-coupled devices, similar to those used in consumer digital cameras, in modern broadband systems. The emitting and detecting devices are often referred to as the optodes. Many systems incorporate 2 or more channels allowing monitoring of multiple tissue regions of interest simultaneously.14
Light Absorption and Scattering
Considering an idealized situation (Fig. 1A), where the only cause of light attenuation between a source and its detector is its absorption by chromophores, the attenuation of light at a given wavelength is described by the Beer-Lambert law. This law states that attenuation is directly proportional to 3 variables: chromophore concentration, the distance traveled by the light between the source and detector, and the absorption coefficient of the chromophore, which describes the absorptive properties of a chromophore at a given wavelength. In this hypothetical situation, chromophore concentration can be accurately calculated using the measured degree of light attenuation by the chromophore in association with knowledge of the source–detector separation and relevant absorption coefficient.
However, biological tissue poses a much more complex situation. Although chromophore concentration is still directly related to light absorption, in the NIR it is light scattering that is the major contributor to attenuation in most biological tissue, including the adult head.15 First, not all emitted light reaches the detector because some of it is scattered away from the detector, giving rise to scattering losses. Second, some of the light that reaches the detector will have been scattered multiple times and therefore have travelled a much greater distance than the actual source–detector distance (Fig. 1B). This means that the detected light path length is significantly greater than the source–detector separation. In the adult head, scattering attenuates NIR light to such an extent that it cannot pass across the whole head; therefore, reflectance spectroscopy, where the light source and detector are placed on adjacent areas of the head, must be used.16
All NIRS techniques rely on a measure of optical attenuation, which is the total loss of light caused by absorption and scattering. Several approaches can be used to derive physiologically relevant signals from measurements of optical attenuation, either by directly measuring optical scattering or accounting for its effects (Table 1).
- Differential spectroscopy using a modification of the Beer-Lambert law is the simplest form of in vivo NIRS, although it is now rarely used in adult clinical practice. This technique assumes that light scattering remains constant during the measurement period, and that measured changes in attenuation are caused only by changes in absorption. Therefore, only changes in chromophore concentration from an arbitrary baseline point (i.e., trends) can be measured.17 The scale of measured changes is dependent on the application of a differential path length factor, which relates the actual mean distance traveled by individual photons to the actual source–detector distance, and which must be defined a priori.18
- Multidistance spectroscopy, also called spatially resolved spectroscopy (SRS), is a technique that is commonly used in commercial cerebral oximeters (Fig. 2A). An array of closely spaced detectors is used to measure light attenuation as a function of source–detector separation and, by combining these measures with an estimation of the wavelength dependency of light scattering, it is possible to derive a scaled absolute hemoglobin concentration, that is, the relative proportions of O2Hb and HHb, from which a tissue oxygen saturation (TOS) can be calculated.19
where [O2Hb] is oxyhemoglobin concentration and [HHb] is deoxyhemoglobin concentration.
- Frequency-resolved (or domain) spectroscopy modulates the intensity of light at a known radio frequency and directly measures the phase shift and degree of light attenuation (Fig. 2B). Models of light transport calculate the absolute values of absorption and scattering, which can be used to approximate absolute chromophore concentration.20
- Time-resolved spectroscopy uses an ultrashort pulse of light, typically a few picoseconds in duration, emitted by a laser (as opposed to light-emitting diode) light source. A histogram of the number of photons detected and their arrival time to the detector is constructed (Fig. 2C). This is called the temporal point spread function and is interpreted with the aid of a model of light transport to calculate the absolute values of absorption and scattering, and thereby the absolute chromophore concentration.21
The accurate derivation of chromophore concentration is tightly linked to the wavelengths of light chosen, and optical absorption at 1 wavelength for each chromophore of interest must be known. The wavelengths of NIR light used in commercial devices are usually selected to be sensitive to hemoglobin; therefore commercial devices generally use wavelengths between 700 nm and 850 nm where the absorption spectra of O2Hb and HHb are maximally separated, and there is minimal overlap with that of water absorption (980 nm).13 Early commercial NIRS systems used 2 wavelengths, limiting their use to the measurement of 2 chromophores, namely, O2Hb and HHb.22 The addition of more wavelengths brings improved accuracy and can be achieved through the use of additional light sources that emit light at discrete wavelengths, or by using broadband spectroscopy systems that make use of a “white” light source emitting a continuous spectrum of light within the NIR. The use of multiple wavelengths is of particular importance when measuring CCO.23
All NIRS methodologies require an algorithm to translate measured changes in light attenuation to a physiologic measure such as changes in O2Hb and HHb concentrations and tissue oxygen saturation. Algorithmic formulae are complex, and examples of nonproprietary algorithms have been described in detail elsewhere.23 Their validity is contingent on certain assumptions; for example, the modified Beer-Lambert law and SRS, the basic methodologies used by many commercial oximeters, assume that optical scattering remains constant throughout the measurements. The central importance of the algorithm is highlighted by studies showing that application of different algorithms to the same optical data yields significantly different chromophore concentrations.23 The variability induced by different algorithms means that there are differences in the cerebral tissue oxygen saturation (SctO2) measured by NIRS-based cerebral oximeters produced by different manufacturers, making comparison between them problematic.24–26
Given the crucial importance of the algorithm in the validity of NIRS-derived data, confidence in a particular device can only be assured when the algorithms on which it relies are open to scrutiny. Some of the devices cited in this review, such as the NIRO series (Hamamatsu Photonics K.K., Hamamatsu City, Japan), use openly published algorithms that are based on models of light transport to calculate SctO2,27 whereas others, such as the INVOS series (Somanetics, Troy, MI), apply proprietary algorithms, or modify algorithms with undeclared scaling factors.9
The variety of approaches to the derivation of chromophore concentration has given rise to multiple NIRS variables and a varied nomenclature. Changes in hemoglobin concentration derived using the modified Beer-Lambert law are usually reported as changes in O2Hb and HHb in micromolar units. Derived hemoglobin indices, such as the total hemoglobin concentration (the sum of O2Hb and HHb) and the hemoglobin difference concentration (the difference between O2Hb and HHb), can also be calculated. Total hemoglobin concentration is often considered to be a surrogate for cerebral blood volume, and the hemoglobin difference concentration for cerebral blood flow.27 Commercial indices include the multiple SctO2 equivalents coined by various manufacturers for their devices. For example, regional cerebral saturation (rsO2) is measured by the INVOS series, tissue oxygenation index by the NIRO series, and SctO2 by the FORE-SIGHT device (CAS-Medical Systems, Brandford, CT). NIR cerebral oximetry does not rely on pulsatile flow but measures a weighted average of arterial, capillary, and venous compartments in proportion to their relative intracranial volumes within the field of view.28 Commercial cerebral oximeters assume a fixed ratio of either 70:30 or 75:25 for venous to arterial blood volume, depending on the manufacturer, and all ignore capillary volume, which is small (approximately 2%). Cerebral oximetry provides real-time information on the balance between cerebral oxygen supply and demand. The total hemoglobin index is a normalized measure of total hemoglobin concentration derived by some SRS devices. CCO is the final electron acceptor in the mitochondrial electron transport chain and is responsible for >95% of oxygen metabolism.29 Its oxidation status also reflects the balance between cerebral energy supply and demand, and NIRS-derived measurement of CCO has been validated as a measure of cellular energy status.30 CCO is a potential biomarker of cellular metabolic state in the clinical setting, but the challenges of measuring CCO using NIRS are substantial. It is present in much lower concentrations in the tissue than O2Hb and HHb and has an absorption spectrum overlapping that of these chromophores. However, multiwavelength NIRS equipment optimized for measurement of CCO in adults has been used successfully to measure changes in the oxidation status of CCO in adults.31,32
Commercial NIR Spectrometers
First-generation clinical monitors used continuous-wave NIRS and the modified Beer-Lambert law and provided only changes in O2Hb and HHb concentrations referenced to an arbitrary baseline.9 The INVOS 3100 (Somanetics, Troy, MI) was the first cerebral oximeter to be approved by the United States Food and Drug Administration and, because it provided an absolute measure of SctO2, its introduction to the market in the early 1990s reinvigorated clinical interest in NIRS. The majority of cerebral oximeters in clinical use today are manufactured by Somanetics (marketed by Covidien, Dublin, Ireland),13 and therefore much of the clinical data reported in this review has been generated using various INVOS devices. Many other cerebral oximeters are also commercially available from multiple manufacturers (Table 1). Some manufacturers are now combining NIRS and other technology into devices with multimodal capability. For example, the CerOx (Ornim Medical Ltd., Lod, Israel) uses a single, noninvasive probe and patented technology (UTLight™, Ornim Medical Ltd.) to provide a brain oximeter and blood flow monitor with a combination of NIR light and ultrasound.13 Prototype devices optimized for specific indications and providing more complex data sets are in widespread use in several research departments, and are likely to inform future clinical applications of NIRS.9 However, there is still no standardization of clinical NIRS devices, and each company continues to develop different optical probes and algorithms (Table 1).
There are several concerns over the clinical application of NIRS, and the one most often highlighted is the potential for “contamination” of the signal by extracranial tissue. Some commercial systems, such as the INVOS, use 2 detectors and a subtraction-based algorithm to deal with this problem. It is assumed that the detecting optode closest to the emitter receives light that has passed mainly through the scalp, whereas that arriving at the farthest detector has mainly passed through brain tissue, although the proprietary algorithms on which this assumption is based are not published. In general terms though, there is weighting in favor of intracerebral tissue with an interoptode spacing (between emitter and detector) >4 cm.33 SRS has high sensitivity and specificity for intracranial changes when appropriate rsO2 thresholds are chosen,34 but is still prone to some degree of extracerebral contamination.35 It has also been suggested that, because mitochondrial density is higher in the brain than in tissues with lower metabolic rates such as the skull and skin, NIRS techniques incorporating CCO measurement may also be less prone to extracerebal contamination.11,12
There is wide intra- and interindividual baseline variability in regional SctO2. The “normal” range lies between 60% and 75%, with a coefficient of variation for absolute baseline values of approximately 10%.25 This means that cerebral oximetry is best used as a trend monitor, and claims of absolute thresholds for cerebral ischemia/hypoxia should be treated with caution (see below). Current commercially available NIRS devices are usually designed to be placed on the forehead and, as with other regional monitoring techniques, it is impossible to detect changes in areas located distant from the monitored site, although global cerebral oxygen sufficiency can be evaluated.11
CLINICAL APPLICATIONS OF NIRS
Although there has been interest in using NIRS to detect cerebral hypoxia/ischemia since its first description, the relatively wide use of the technology in the research setting has not been matched by clinical applications.9 There are many reasons for this, including limitations inherent to the technique, ambiguity regarding what is actually measured, deficiencies in algorithms, difficult translation of instruments designed for research use to the bedside, and data that are not reproducible and specific enough for clinical decision making. In assessing the evidence for the clinical application of NIRS, 4 questions are therefore relevant:
- What is the nature and value of NIRS thresholds that predict cerebral hypoxia/ischemia?
- Does NIRS have any advantages over existing methods for assessing cerebral hypoxia/ischemia?
- Do NIRS-derived measurements carry any prognostic value for the occurrence of cerebral hypoxic/ischemic complications?
- Do NIRS-directed treatment strategies modulate the outcome?
These considerations will inform the following evidence-based review of the clinical use of NIRS. For clarity, changes in SctO2 and its equivalents will be referred to in 2 ways: as a percentage point reduction, or as a percentage reduction from baseline. Thus, an SctO2 decrease from a baseline value of 50% to 40% is expressed as a reduction of 10 percentage points, or a reduction to 80% of baseline values.
Carotid endarterectomy (CEA) is primarily performed for the prevention of embolic/ischemic stroke in patients with atherosclerotic carotid stenosis, yet paradoxically CEA itself carries more than a 2% risk of stroke because of emboli or carotid cross-clamp–related ischemia.36 Strategies to maintain adequate cerebral perfusion and prevent cerebral hypoxic/ischemic injury include intracarotid shunt placement, induced hypertension, and administration of supplemental oxygen. Because these all carry risks of varying degrees, it is desirable to limit their use to patients with evidence of critical hypoxia/ischemia.37
Several methods have been used to assess the adequacy of cerebral oxygen delivery during the cross-clamp period and to inform the critical decision regarding shunt placement. When CEA is performed under regional anesthesia, alteration of mentation is the best monitor of impending ischemia and an indication for shunt placement.38 However, when general anesthesia is required, surrogate measures of cerebral ischemia, including electroencephalography, transcranial Doppler ultrasonography (TCD), somatosensory-evoked potentials (SEPs), and carotid stump pressure have been used.39–41 NIRS, in particular cerebral oximetry using SRS, offers an advantage compared with other modalities in terms of simplicity, and its usefulness during CEA has been the subject of a systematic review.42
The first comparison between NIRS and the established “gold standard” of clinical neurologic examination was published by Samra et al.43 in 2000. In a retrospective analysis of 94 cases of CEA performed under regional anesthesia, 10 patients who developed neurologic symptoms of cerebral ischemia had a significantly greater (mean ± SD) reduction in rsO2 from baseline compared with those with no neurologic signs (63.2% ± 8.4% to 51.0% ± 11.6% and 65.8% ± 8.5% to 61.0% + 9.3%, respectively, P = 0.0002). Using logistic probability analysis, the authors suggested that a reduction in rsO2 of 20 percentage points from baseline provides the optimum sensitivity–specificity balance, with a sensitivity and specificity of 80% and 82.2%, respectively. However, a subsequent study of 594 patients undergoing CEA under general anesthesia used receiver operating characteristic analysis and identified a much smaller reduction of 11.7 percentage points from baseline as the optimal predictor of postoperative neurologic dysfunction.44 In 48 patients undergoing CEA under regional anesthesia, Moritz et al.41 showed that an absolute rsO2 threshold of 59% was 100% sensitive and 47% specific for clinical ischemia defined by changes in neurologic examination. An elegant study by Al-Rawi and Kirkpatrick45 attempted to quantify NIRS-defined carotid cross-clamp–related cerebral ischemia during carotid surgery under general anesthesia. Using an NIRO-300 device, no patient with a reduction in tissue oxygenation index <13 percentage points developed electroencephalographic evidence of ischemia, with a specificity of 93%. Regional and general anesthesia have different effects on the cerebral vasculature, and therefore it seems imprudent to apply ischemic “thresholds” identified during regional anesthesia to procedures performed under general anesthesia, and vice versa.
There are somewhat more consistent data demonstrating agreement between NIRS and other monitoring modalities. In one study, receiver operating characteristic analysis of the best fit of NIRS oximetry, TCD, SEPs, and carotid stump pressure showed largely similar sensitivity and specificity compared with neurologic examination during CEA under regional anesthesia,41 and studies performed during general anesthesia have shown similar levels of agreement.39,40,46 However, the conclusions of these studies are hampered by the selection of potentially ill-founded thresholds for the identification of ischemia, by the interrogation of different regions of the brain by each modality, and by the relatively low specificity of “standard” modalities in predicting neurologic morbidity after CEA.47
One advantage of cerebral oximetry is its high temporal resolution compared with modalities such as SEPs, allowing early monitor-guided modification of surgical and anesthesia management to prevent or minimize cerebral ischemia. It is also theoretically possible to minimize the influence of extracerebral tissue on NIRS signals during CEA by establishing a new baseline value for rsO2 after the external carotid artery has been cross-clamped. However, such “recalibration” is not widely practiced, and the increasing application of techniques such as SRS makes it unnecessary.
In summary, although NIRS is not proven to be superior to other monitoring modalities for identifying critical cerebral ischemia during CEA, the body of evidence suggests broad equivalence, albeit with uncertainty as to the exact NIRS-derived threshold for the identification of critical ischemia. However, the relative ease of use, lack of operator dependence, and temporal resolution of NIRS make it an attractive option. Looking forward, the use of NIRS to guide the manipulation of systemic physiology to minimize the risk of cerebral hypoxia/ischemia during CEA is an area where NIRS might prove to be superior to other modalities.48–50
Poor neurologic outcome remains a concern in patients undergoing cardiac surgery and cardiopulmonary bypass. Stroke occurs in 1% to 3% of patients, but this is overshadowed by the development of long-standing postoperative cognitive dysfunction in >50% of patients.51 Likely mechanisms of injury include emboli and cerebral hypoperfusion, and several different management protocols have been introduced to minimize these risks. However, it is the application of NIRS-guided “brain protection” protocols aimed at optimizing cerebral oxygen delivery during cardiopulmonary bypass that has generated substantial interest.52,53
Early retrospective studies investigating therapeutic strategies aimed at maintaining rsO2 at (various) predetermined levels during cardiac surgery created great optimism that their implementation would reduce the incidence of stroke and postoperative cognitive dysfunction.54,55 However, these encouraging results have not been reproduced in subsequent prospective studies. Slater et al.56 observed that a strategy designed to maintain rsO2 (measured by an INVOS oximeter) in 265 patients undergoing coronary artery bypass grafting had no effect on the incidence of postoperative cognitive decline, which was assessed by a comprehensive battery of neurocognitive and neuropsychological assessments and was found to be 58% and 61% in the intervention and control groups, respectively. A post hoc sensitivity analysis identified a desaturation dose of 3000 second.% (where the dose was defined as the product of length of time and depth of rsO2 <50%) to be an independent predictor of postoperative cognitive decline (odds ratio 2.22, P = 0.048) and extended hospital length of stay (odds ratio 2.71, P = 0.007). The intervention was not successful in restoring rsO2 in all patients. However, because treatment was initiated to reverse desaturation once an rsO2 threshold had been reached, it is possible that any potential outcome benefits are derived from the rapid reversal of cerebral desaturation. This raises the possibility of the existence of a “viability-time threshold,” whereby a time-dependent exposure to ischemia leads to functional impairment, a concept supported by data from animal studies.57 Another potential application of NIRS-guided brain protection during cardiac surgery is its ability to guide arterial blood pressure management. In particular, NIRS has been used to identify the lower limit of cerebral autoregulation to optimize individual blood pressure management during cardiopulmonary bypass.58
Murkin et al59 investigated the effectiveness of a strategy to maintain rsO2 (measured using an INVOS oximeter) at 75% of preoperative values and found no significant reduction in stroke incidence or mortality after cardiac surgery. However, significantly more patients in the control (no intervention) group died or developed major organ morbidity than those in the intervention group, and this was associated with lower baseline and mean rsO2, more episodes of intraoperative cerebral desaturation, and longer intensive care unit and hospital length of stay. Thus, although the overall ability of cerebral oximetry to affect neurologic outcome is unproven, possibly because studies are not sufficiently powered to detect subtle outcome effects, the study by Murkin et al. raises the possibility that the brain might be a surrogate for monitoring the adequacy of other organ perfusion in the absence of direct feedback from those organ systems. The accessibility of the brain to interrogation by NIRS makes it an attractive index organ in this regard. It has been proposed that this principle might be extended to the application of NIRS during the preoperative assessment of other organ function. In a prospective observational study of 1178 patients undergoing cardiac surgery with cardiopulmonary bypass, low preoperative rsO2 was significantly correlated with postoperative biomarkers of cardiac and renal injury, and with cardiac dysfunction.60 rsO2 was lower in those who had died by 30 days compared with those who survived (median [95% confidence interval] 58% [50.7%–62.0%] vs 64% [64%–65%], respectively, P < 0.0001), and preoperative rsO2 < 50% was an independent risk factor for 30-day and 1-year mortality. The authors suggest that using preoperative rsO2 as a surrogate of other organ perfusion might be a useful addition to the risk stratification of patients undergoing cardiopulmonary bypass.
Rather than relying on hemoglobin variables, NIRS-defined tissue viability-time thresholds might be better quantified by the measurement of CCO, because of its crucial role in mitochondrial oxidative energy metabolism.11 In an observational study of 66 patients undergoing aortic arch surgery, Kakihana et al.61 found that patients who developed a persistent and prolonged reduction in CCO during cardiopulmonary bypass had a higher incidence of neurologic complications compared with those in whom a transient, or no, change in CCO oxidation state was observed. Furthermore, the potential benefits of NIRS monitoring during high-risk cardiac surgery might be enhanced by the application of novel interpretative techniques that use mathematical models of NIRS variables to quantify time limits for cerebral hypoxia/ischemia.62
Despite the absence of compelling data to support the use of NIRS-guided management strategies to reduce the incidence of postoperative cognitive dysfunction and stroke, NIRS is being increasingly used to monitor and manage cerebral oxygenation during cardiac surgery because of its lack of use-associated risk and modest cost.53,63 However, a recent review suggesting that the neurocognitive decline after cardiac bypass surgery may not only be related to the intervention but might also reflect the natural decline of patients with multiple comorbidities raises an important question about the role of neuromonitoring, including NIRS, to guide treatment during cardiopulmonary bypass.64
Routine NIRS Monitoring During General Anesthesia
It has been suggested that NIRS might be useful for monitoring the healthy but at-risk brain during routine surgical procedures under general anesthesia. Although the potential to continuously monitor cerebral well-being is an attractive proposition, and early detection of cerebral desaturation might lead to targeted intervention that could improve perioperative outcome, the evidence for such benefit has thus far proved elusive.65
A cohort study of 60 patients aged >65 years found that cerebral desaturation (defined as rsO2 decline to 75% of baseline, or to 80% of baseline if initial values were <50%), measured with an INVOS oximeter, occurred in 26% of patients undergoing scheduled major nonvascular abdominal surgery.66 Baseline, minimal, and mean rsO2 values were significantly and negatively correlated, and the length of time with an absolute rsO2 value <50% positively correlated, with hospital length of stay. In a randomized-controlled study of 122 otherwise healthy elderly patients, also undergoing nonvascular abdominal surgery, an intervention group, in which rsO2 was maintained ≥75% of baseline, was compared with a control group, in which anesthesia was managed routinely and rsO2 was monitored but not displayed.67 Although the overall mean rsO2 was unsurprisingly higher in patients in the intervention than in the control group (66% vs 61%, respectively, P = 0.002), there was no difference in the incidence of cerebral desaturation in the 2 groups: rsO2 decreased below 75% of baseline in 20% and 23% of patients in the intervention and control groups, respectively (P = 0.82). However, control patients who developed intraoperative cerebral desaturation had a lower mini mental state examination score on the 7th postoperative day and longer hospital length of stay than patients in the intervention group (P = 0.02). The failure of the intervention to prevent cerebral desaturation in 20% of patients merits further consideration if such protocols are to be introduced more widely.
An area of emerging interest is the application of cerebral oximetry in patients undergoing surgery in the beach chair position. Episodes of severe hypotension occur in up to 20% of patients,68 and ischemia-related cerebrovascular events have been reported,69 suggesting that quantitative monitoring of impending cerebral ischemia would be desirable in patients undergoing surgery in the steep head-up position. There have been isolated case reports describing the use of cerebral oximetry for surgery in the beach chair position.70,71 An observational study of 124 patients undergoing shoulder arthroscopy in beach chair and lateral decubitus positions used a FORE-SIGHT oximeter and demonstrated cerebral desaturation (>20% decrease in SctO2 from baseline or SctO2 <55% for >15 seconds) in 80.3% of patients in the beach chair compared with none in the lateral decubitus position.72 SctO2 was significantly lower in the beach chair than in the lateral decubitus position as was the median number of cerebral desaturation events per subject (4, range 0–38 vs 0, range 0–0, respectively, P < 0.0001). Despite this apparently alarming incidence of cerebral desaturation in the beach chair position, there were no postoperative neurologic abnormalities. However, desaturation was associated with an increased risk of postoperative nausea and vomiting, and it is possible that this was related to gastrointestinal hypoperfusion in those who experienced intraoperative cerebral desaturation, supporting the hypothesis that brain oxygenation might be a surrogate of the adequacy of non-neurologic organ perfusion. This study raises several important issues. First, it is not clear how the observed incidence of cerebral desaturation relates to the exceedingly low incidence of postoperative neurologic damage, estimated to be 3 of 100,000,69 in the large numbers of patients who undergo surgery in the beach chair position. Second, it is not clear what the episodes of cerebral desaturation actually represent. In this study, cardiorespiratory variables were unaffected by the operative position, and the cerebral metabolic rate is unlikely to have changed substantially during the study period. Therefore, the balance between cerebral oxygen supply and demand is also unlikely to have changed and SctO2 would therefore have been expected to be stable. Changes in intracranial geometry and cerebral arteriovenous ratio occur during movement from supine to upright position and we speculate that this might account, at least in part, for the changes in measured SctO2 in this and similar studies.65 Although the devastating effects of ischemic cerebral complications of surgery in the beach chair position must be avoided, the role of NIRS oximetry in monitoring and preventing them is not clear.
In acute brain injury, cellular hypoxia/ischemia is a key component of the multiple pathophysiological processes that contribute to secondary injury. Because cerebral oxygenation, hemodynamic, and metabolic variables are attractive targets for monitor-guided therapeutic intervention,1 a logical application for NIRS is after acute brain injury, where secondary ischemic injury is common and is associated with adverse outcomes. However, there has been limited investigation of the usefulness of NIRS in this area, and there have been no outcome studies. Investigations in the adult neurocritical care population have been solely observational and serve to highlight 2 key difficulties in investigating NIRS-derived variables in this context: (1) the definition of ischemic thresholds in an injured brain with acutely disordered hemodynamic and metabolic functions, and (2) the lack of a gold standard against which to compare NIRS-derived measures.12 Factors such as the presence of intracranial hematoma, cerebral edema, and subarachnoid blood present further challenges because they may invalidate some of the assumptions on which NIRS algorithms are based. In fact, this has been used to advantage in studies using NIRS to identify intracranial hematomas73 and cerebral edema.74
A small observational study of 18 patients with traumatic brain injury (TBI) identified an association between increasing length of time with rsO2 values <60% (measured with an INVOS oximeter) and mortality, intracranial hypertension, and compromised cerebral perfusion pressure.75 A more recent study compared brain tissue oxygen tension (PbtO2) with rsO2 in 22 patients over a 16-hour period after severe TBI.76 An rsO2 <60% was moderately accurate for the prediction of “severe” brain hypoxia (PbtO2 <12 mm Hg), but was poor at detecting “moderate” hypoxia (PbtO2, 12–15 mm Hg) with sensitivities and specificities of 73% and 86%, and 62% and 49%, respectively. The authors conclude that rsO2 is an inadequate substitute for PbtO2 for routine monitoring of cerebral oxygenation and, although this is reasonable, it fails to consider the different physiological variables and brain regions being monitored by the 2 techniques.
There has been recent interest in using NIRS to monitor cortical changes during cerebral vasospasm after aneurysmal subarachnoid hemorrhage (SAH). In one study of 32 patients undergoing coil embolization, the rate of decline of rsO2 (measured using an INVOS oximeter) was 3.5%.minutes-1 greater in patients who developed vasospasm during the procedure than in those who did not.77 Yokose et al.78 used time-resolved spectroscopy (TRS-20, Hamamatsu Photonics) to demonstrate that vasospasm could be predicted with 100% sensitivity and 85.7% specificity in 14 patients with predominantly poor-grade SAH, and that a threshold of a 3.9% to 6.4% decline in cortical oxygen saturation was optimal for the identification of ischemia. In this study, the reliability of repeated NIRS measurements over time was ensured by the use of computed tomography image guidance to position the NIRS optodes. This novel approach allowed consistent measurement of the same cortical area on consecutive days. The introduction of time-resolved devices to the clinical setting brings the prospect of the measurement of multiple chromophore concentrations in absolute terms at the bedside.
Measurement of CCO concentration may provide additional information about metabolic failure, and aid in the determination of ischemic thresholds after brain injury.12 Using a customized broadband spectroscopy system, Tisdall et al.31,32 found that changes in CCO concentration correlated with measures of cerebral oxygen delivery in healthy volunteers and subsequently demonstrated an increase in CCO concentration during normobaric hyperoxia in a pilot study of 8 patients after TBI. The lack of a gold standard against which to compare CCO concentration remains an obstacle to its interpretation. However, the development of mathematical models of brain hemodynamics and metabolism allows in silico derivation of physiologic variables, which can be compared with measured signals and thereby facilitate their interpretation.79
Impairment of cerebrovascular autoregulation renders the brain more susceptible to ischemic insults and is thus associated with poor outcome after brain injury. NIRS has recently been used to monitor cerebrovascular reactivity in a variety of clinical settings and, although the therapeutic value of monitoring and managing autoregulation is not clear, this has become an area of intense investigation. In a study of 40 patients with severe TBI, Zweifel et al.80 identified a statistically significant correlation between a noninvasive NIRS hemoglobin volume–based measure of cerebrovascular reactivity, and arterial blood pressure and PRx, an index of pressure reactivity derived from invasive intracranial and arterial pressures. The same group also demonstrated a correlation between an oxygen-based NIRS-derived measure of cerebrovascular reactivity and Mx, a TCD index derived from middle cerebral artery blood flow velocity in a group of 27 patients after poor-grade SAH.81 It is suggested that NIRS-derived measures of cerebrovascular autoregulation may be able to guide optimization of cerebral hemodynamics, including cerebral perfusion pressure, although evidence of its clinical usefulness is currently sparse.
The complex and nonlinear relationships between NIRS and other variables, such as intracranial pressure and blood flow velocity, that are conventionally used to assess cerebrovascular reactivity makes it necessary to apply more complex analytical techniques. For example, wavelet-based techniques aid the interpretation of complex time variant signals because they focus analysis to specific features of interest within the time and frequency domains simultaneously, producing qualitative and quantitative evidence of cerebrovascular reactivity that is not available from other methods.82 In this way, NIRS might provide a sensitive means of monitoring cerebral autoregulation that would readily translate into clinical practice. Furthermore, the noninvasive nature of this technique allows it to be applied in a broad range of clinical neuroscience scenarios, including those outside the intensive care unit.
Developing Applications and Future Directions
Aside from identifying cerebral ischemia/hypoxia, NIRS can also characterize the cerebral hemodynamic and metabolic responses associated with neuronal function and dysfunction. There is extensive literature on functional brain monitoring using NIRS, and this has been reviewed in detail.7 Functional brain activation has also been used to identify pathological states. For example, NIRS-measured hemodynamic responses to functional activation are either absent or abnormally lateralized in patients with Alzheimer’s disease.83,84 NIRS-derived measures of cerebral blood volume have also shown potential in the diagnosis of seizure type,85 focus localization,86 and lateralization of language areas in the preoperative assessment of patients for epilepsy surgery.87 There is also interest in the application of various NIRS techniques to assess the dynamic pathophysiological processes, including changes in cerebral perfusion, that occur after acute ischemic stroke.88
The emergence of new technologies will be crucial in broadening the applications of NIRS. Notable developments include the use of supercontinuum light sources, which allow the combination of time-resolved and broadband spectroscopy to yield absolute measurements of optical absorption and scattering across a range of wavelengths, and thus of multiple chromophores.89 Diffuse correlation spectroscopy, which has the potential to provide noninvasive measures of cerebral blood flow in addition to hemoglobin oxygen saturation, is also likely to evolve into a clinical technique.90 Combinations of developing and established technologies will allow the quantification of variables of greater physiological and clinical relevance than existing systems, including the possibility for in vivo molecular imaging.22
Although NIRS has potential as a noninvasive brain monitor, the story thus far is one of incompletely realized potential. While there is an increasing body of evidence suggesting that treatment paradigms directed by NIRS-based cerebral oximeters may lead to improved outcome after cardiac surgery, there are no data to support the routine application of NIRS during anesthesia and surgery more widely. There is also no evidence to support its use in brain injury where it might be expected to have a key role. Although increasingly sophisticated apparatuses, including broadband and time-resolved systems, provide tantalizing insights into the potential of NIRS to measure multiple chromophores in the research setting, these innovations have yet to translate into clinical practice. NIRS has many advantages over other neuromonitoring techniques; it is user-friendly, noninvasive, and is able to make measurements over multiple regions of interest simultaneous with high temporal resolution. NIRS-based techniques have real potential to monitor regional cerebral oxygenation, hemodynamics, and metabolism, and guide therapeutic brain protection strategies, but further investigation and technological advances are necessary before they can be introduced more widely into clinical practice.
Name: Arnab Ghosh, MBChB, BSc (Hons), MRCS.
Contribution: This author helped write the manuscript.
Attestation: Arnab Ghosh approved the final manuscript.
Name: Clare Elwell, PhD.
Contribution: This author helped write the manuscript.
Attestation: Clare Elwell approved the final manuscript.
Name: Martin Smith, MBBS, FRCA, FFICM.
Contribution: This author helped write the manuscript
Attestation: Martin Smith approved the final manuscript.
This manuscript was handled by: Gregory J. Crosby, MD.
1. Tisdall MM, Smith M. Multimodal monitoring in traumatic brain injury: current status and future directions. Br J Anaesth. 2007;99:61–7
2. Highton D, Elwell C, Smith M. Noninvasive cerebral oximetry: is there light at the end of the tunnel? Curr Opin Anaesthesiol. 2010;23:576–81
3. Brazy JE, Lewis DV, Mitnick MH, Jöbsis vander Vliet FF. Noninvasive monitoring of cerebral oxygenation in preterm infants: preliminary observations. Pediatrics. 1985;75:217–25
4. Lloyd-Fox S, Blasi A, Elwell CE. Illuminating the developing brain: the past, present and future of functional near infrared spectroscopy. Neurosci Biobehav Rev. 2010;34:269–84
5. van Bel F, Lemmers P, Naulaers G. Monitoring neonatal regional cerebral oxygen saturation in clinical practice: value and pitfalls. Neonatology. 2008;94:237–44
6. Ferrari M, Muthalib M, Quaresima V. The use of near-infrared spectroscopy in understanding skeletal muscle physiology: recent developments. Philos Transact A Math Phys Eng Sci. 2011;369:4577–90
7. Hoshi Y. Functional near-infrared spectroscopy: current status and future prospects. J Biomed Opt. 2007;12:062106
8. Rolfe P. In vivo
near-infrared spectroscopy. Annu Rev Biomed Eng. 2000;2:715–54
9. Wolf M, Ferrari M, Quaresima V. Progress of near-infrared spectroscopy and topography for brain and muscle clinical applications. J Biomed Opt. 2007;12:062104
10. Jöbsis FF. Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science. 1977;198:1264–7
11. Kakihana Y, Matsunaga A, Yasuda T, Imabayashi T, Kanmura Y, Tamura M. Brain oxymetry in the operating room: current status and future directions with particular regard to cytochrome oxidase. J Biomed Opt. 2008;13:033001
12. Smith M, Elwell C. Near-infrared spectroscopy: shedding light on the injured brain. Anesth Analg. 2009;108:1055–7
13. Ferrari M, Quaresima V. Near infrared brain and muscle oximetry: from discovery to current applications J Near Infrared Spectrosc. 2012;20:1–14
14. Ferrari M, Mottola L, Quaresima V. Principles, techniques, and limitations of near infrared spectroscopy. Can J Appl Physiol. 2004;29:463–87
15. Cheong WF, Prahl SA, Welch AJ. A review of the optical properties of biological tissues IEEE J. Quantum Electron. 1990;26:2166–85
16. Okada E, Delpy DT. Near-infrared light propagation in an adult head model. II. Effect of superficial tissue thickness on the sensitivity of the near-infrared spectroscopy signal. Appl Opt. 2003;42:2915–22
17. Delpy DT, Cope M, van der Zee P, Arridge S, Wray S, Wyatt J. Estimation of optical pathlength through tissue from direct time of flight measurement. Phys Med Biol. 1988;33:1433–42
18. Duncan A, Meek JH, Clemence M, Elwell CE, Tyszczuk L, Cope M, Delpy DT. Optical pathlength measurements on adult head, calf and forearm and the head of the newborn infant using phase resolved optical spectroscopy. Phys Med Biol. 1995;40:295–304
19. Suzuki S, Takasaki S, Ozaki T, Kobayashi Y. A tissue oxygenation monitor using NIR spatially resolved spectroscopy Proc SPIE. 1999;3597:592
20. Fantini S, Franceschini MTuchin VV ed. Frequency-domain techniques for tissue spectroscopy and imaging. In: Handbook of Optical Biomedical Diagnostics.. 2002 Bellingham, WA SPIE:405–43
21. Wabnitz H, Moeller M, Liebert A, Obrig H, Steinbrink J, Macdonald R. Time-resolved near-infrared spectroscopy and imaging of the adult human brain. Adv Exp Med Biol. 2010;662:143–8
22. Elwell CE, Cooper CE. Making light work: illuminating the future of biomedical optics. Philos Transact A Math Phys Eng Sci. 2011;369:4358–79
23. Matcher SJ, Elwell CE, Cooper CE, Cope M, Delpy DT. Performance comparison of several published tissue near-infrared spectroscopy algorithms. Anal Biochem. 1995;227:54–68
24. Gagnon RE, Macnab AJ, Gagnon FA, Blackstock D, LeBlanc JG. Comparison of two spatially resolved NIRS oxygenation indices. J Clin Monit Comput. 2002;17:385–91
25. Thavasothy M, Broadhead M, Elwell C, Peters M, Smith M. A comparison of cerebral oxygenation as measured by the NIRO 300 and the INVOS 5100 Near-Infrared Spectrophotometers. Anaesthesia. 2002;57:999–1006
26. Yoshitani K, Kawaguchi M, Tatsumi K, Kitaguchi K, Furuya H. A comparison of the INVOS 4100 and the NIRO 300 near-infrared spectrophotometers. Anesth Analg. 2002;94:586–90
27. Elwell CE A Practical Users Guide to Near Infrared Spectroscopy. 1995 Hamamatsu, Japan Hamamatsu Photonics KK,
28. Ito H, Ibaraki M, Kanno I, Fukuda H, Miura S. Changes in the arterial fraction of human cerebral blood volume during hypercapnia and hypocapnia measured by positron emission tomography. J Cereb Blood Flow Metab. 2005;25:852–7
29. Richter OM, Ludwig B. Cytochrome c oxidase–structure, function, and physiology of a redox-driven molecular machine. Rev Physiol Biochem Pharmacol. 2003;147:47–74
30. Springett RJ, Wylezinska M, Cady EB, Hollis V, Cope M, Delpy DT. The oxygen dependency of cerebral oxidative metabolism in the newborn piglet studied with 31P NMRS and NIRS. Adv Exp Med Biol. 2003;530:555–63
31. Tisdall MM, Tachtsidis I, Leung TS, Elwell CE, Smith M. Near-infrared spectroscopic quantification of changes in the concentration of oxidized cytochrome c oxidase in the healthy human brain during hypoxemia. J Biomed Opt. 2007;12:024002
32. Tisdall MM, Tachtsidis I, Leung TS, Elwell CE, Smith M. Increase in cerebral aerobic metabolism by normobaric hyperoxia after traumatic brain injury. J Neurosurg. 2008;109:424–32
33. Germon TJ, Evans PD, Barnett NJ, Wall P, Manara AR, Nelson RJ. Cerebral near infrared spectroscopy: emitter-detector separation must be increased. Br J Anaesth. 1999;82:831–7
34. 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;32:2492–500
35. Davie SN, Grocott HP. Impact of extracranial contamination on regional cerebral oxygen saturation: a comparison of three cerebral oximetry technologies. Anesthesiology. 2012;116:834–40
36. Brott TG, Hobson RW 2nd, Howard G, Roubin GS, Clark WM, Brooks W, Mackey A, Hill MD, Leimgruber PP, Sheffet AJ, Howard VJ, Moore WS, Voeks JH, Hopkins LN, Cutlip DE, Cohen DJ, Popma JJ, Ferguson RD, Cohen SN, Blackshear JL, Silver FL, Mohr JP, Lal BK, Meschia JFCREST Investigators. . Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med. 2010;363:11–23
37. Howell SJ. Carotid endarterectomy. Br J Anaesth. 2007;99:119–31
38. Evans WE, Hayes JP, Waltke EA, Vermilion BD. Optimal cerebral monitoring during carotid endarterectomy: neurologic response under local anesthesia. J Vasc Surg. 1985;2:775–7
39. Friedell ML, Clark JM, Graham DA, Isley MR, Zhang XF. Cerebral oximetry does not correlate with electroencephalography and somatosensory evoked potentials in determining the need for shunting during carotid endarterectomy. J Vasc Surg. 2008;48:601–6
40. Grubhofer G, Plöchl W, Skolka M, Czerny M, Ehrlich M, Lassnigg A. Comparing Doppler ultrasonography and cerebral oximetry as indicators for shunting in carotid endarterectomy. Anesth Analg. 2000;91:1339–44
41. Moritz S, Kasprzak P, Arlt M, Taeger K, Metz C. Accuracy of cerebral monitoring in detecting cerebral ischemia during carotid endarterectomy: a comparison of transcranial Doppler sonography, near-infrared spectroscopy, stump pressure, and somatosensory evoked potentials. Anesthesiology. 2007;107:563–9
42. Pennekamp CW, Bots ML, Kappelle LJ, Moll FL, de Borst GJ. The value of near-infrared spectroscopy measured cerebral oximetry during carotid endarterectomy in perioperative stroke prevention: a review. Eur J Vasc Endovasc Surg. 2009;38:539–45
43. Samra SK, Dy EA, Welch K, Dorje P, Zelenock GB, Stanley JC. Evaluation of a cerebral oximeter as a monitor of cerebral ischemia during carotid endarterectomy. Anesthesiology. 2000;93:964–70
44. Mille T, Tachimiri ME, Klersy C, Ticozzelli G, Bellinzona G, Blangetti I, Pirrelli S, Lovotti M, Odero A. Near infrared spectroscopy monitoring during carotid endarterectomy: which threshold value is critical? Eur J Vasc Endovasc Surg. 2004;27:646–50
45. Al-Rawi PG, Kirkpatrick PJ. Tissue oxygen index: thresholds for cerebral ischemia using near-infrared spectroscopy. Stroke. 2006;37:2720–5
46. Manwaring ML, Durham CA, McNally MM, Agle SC, Parker FM, Stoner MC. Correlation of cerebral oximetry with internal carotid artery stump pressures in carotid endarterectomy. Vasc Endovascular Surg. 2010;44:252–6
47. Nemoto EM. No absolutes in neuromonitoring for carotid endarterectomy. Stroke. 1999;30:895
48. Giustiniano E, Alfano A, Battistini GM, Gavazzeni V, Spoto MR, Cancellieri F. Cerebral oximetry during carotid clamping: is blood pressure raising necessary? J Cardiovasc Med (Hagerstown). 2010;11:522–8
49. Picton P, Chambers J, Shanks A, Dorje P. The influence of inspired oxygen fraction and end-tidal carbon dioxide on post-cross-clamp cerebral oxygenation during carotid endarterectomy under general anesthesia. Anesth Analg. 2010;110:581–7
50. Stoneham MD, Lodi O, de Beer TC, Sear JW. Increased oxygen administration improves cerebral oxygenation in patients undergoing awake carotid surgery. Anesth Analg. 2008;107:1670–5
51. Newman MF, Mathew JP, Grocott HP, Mackensen GB, Monk T, Welsh-Bohmer KA, Blumenthal JA, Laskowitz DT, Mark DB. Central nervous system injury associated with cardiac surgery. Lancet. 2006;368:694–703
52. Fedorow C, Grocott HP. Cerebral monitoring to optimize outcomes after cardiac surgery. Curr Opin Anaesthesiol. 2010;23:89–94
53. Vohra HA, Modi A, Ohri SK. Does use of intra-operative cerebral regional oxygen saturation monitoring during cardiac surgery lead to improved clinical outcomes? Interact Cardiovasc Thorac Surg. 2009;9:318–22
54. Edmonds HL Jr. Protective effect of neuromonitoring during cardiac surgery. Ann N Y Acad Sci. 2005;1053:12–9
55. Goldman S, Sutter F, Ferdinand F, Trace C. Optimizing intraoperative cerebral oxygen delivery using noninvasive cerebral oximetry decreases the incidence of stroke for cardiac surgical patients. Heart Surg Forum. 2004;7:E376–81
56. Slater JP, Guarino T, Stack J, Vinod K, Bustami RT, Brown JM 3rd, Rodriguez AL, Magovern CJ, Zaubler T, Freundlich K, Parr GV. Cerebral oxygen desaturation predicts cognitive decline and longer hospital stay after cardiac surgery. Ann Thorac Surg. 2009;87:36–44
57. Kurth CD, McCann JC, Wu J, Miles L, Loepke AW. Cerebral oxygen saturation-time threshold for hypoxic-ischemic injury in piglets. Anesth Analg. 2009;108:1268–77
58. Joshi B, Ono M, Brown C, Brady K, Easley RB, Yenokyan G, Gottesman RF, Hogue CW. Predicting the limits of cerebral autoregulation during cardiopulmonary bypass. Anesth Analg. 2012;114:503–10
59. Murkin JM, Adams SJ, Novick RJ, Quantz M, Bainbridge D, Iglesias I, Cleland A, Schaefer B, Irwin B, Fox S. Monitoring brain oxygen saturation during coronary bypass surgery: a randomized, prospective study. Anesth Analg. 2007;104:51–8
60. Heringlake M, Garbers C, Käbler JH, Anderson I, Heinze H, Schön J, Berger KU, Dibbelt L, Sievers HH, Hanke T. Preoperative cerebral oxygen saturation and clinical outcomes in cardiac surgery. Anesthesiology. 2011;114:58–69
61. Kakihana Y, Matsunaga A, Tobo K, Isowaki S, Kawakami M, Tsuneyoshi I, Kanmura Y, Tamura M. Redox behavior of cytochrome oxidase and neurological prognosis in 66 patients who underwent thoracic aortic surgery. Eur J Cardiothorac Surg. 2002;21:434–9
62. Fischer GW, Benni PB, Lin HM, Satyapriya A, Afonso A, Di Luozzo G, Griepp RB, Reich DL. Mathematical model for describing cerebral oxygen desaturation in patients undergoing deep hypothermic circulatory arrest. Br J Anaesth. 2010;104:59–66
63. Murkin JM, Arango M. Near-infrared spectroscopy as an index of brain and tissue oxygenation. Br J Anaesth. 2009;103 Suppl 1:i3–13
64. Selnes OA, Gottesman RF, Grega MA, Baumgartner WA, Zeger SL, McKhann GM. Cognitive and neurologic outcomes after coronary-artery bypass surgery. N Engl J Med. 2012;366:250–7
65. Smith M. Shedding light on the adult brain: a review of the clinical applications of near-infrared spectroscopy. Philos Transact A Math Phys Eng Sci. 2011;369:4452–69
66. Casati A, Fanelli G, Pietropaoli P, Proietti R, Tufano R, Montanini S, Danelli G, Nuzzi M, Mentegazzi F, Torri G, Martani C, Spreafico E, Fierro G, Pugliese F, De Cosmo G, Aceto P, Servillo G, Monaco FCollaborative Italian Study Group on Anaesthesia in Elderly Patients. . Monitoring cerebral oxygen saturation in elderly patients undergoing general abdominal surgery: a prospective cohort study. Eur J Anaesthesiol. 2007;24:59–65
67. Casati A, Fanelli G, Pietropaoli P, Proietti R, Tufano R, Danelli G, Fierro G, Fierro G, De Cosmo G, Servillo G. Continuous monitoring of cerebral oxygen saturation in elderly patients undergoing major abdominal surgery minimizes brain exposure to potential hypoxia. Anesth Analg. 2005;101:740–7
68. D’Alessio JG, Rosenblum M, Shea KP, Freitas DG. A retrospective comparison of interscalene block and general anesthesia for ambulatory surgery shoulder arthroscopy. Reg Anesth. 1995;20:62–8
69. Friedman DJ, Parnes NZ, Zimmer Z, Higgins LD, Warner JJ. Prevalence of cerebrovascular events during shoulder surgery and association with patient position. Orthopedics. 2009;32:256
70. Dippmann C, Winge S, Nielsen HB. Severe cerebral desaturation during shoulder arthroscopy in the beach-chair position. Arthroscopy. 2010;26:S148–50
71. Fischer GW, Torrillo TM, Weiner MM, Rosenblatt MA. The use of cerebral oximetry as a monitor of the adequacy of cerebral perfusion in a patient undergoing shoulder surgery in the beach chair position. Pain Pract. 2009;9:304–7
72. Murphy GS, Szokol JW, Marymont JH, Greenberg SB, Avram MJ, Vender JS, Vaughn J, Nisman M. Cerebral oxygen desaturation events assessed by near-infrared spectroscopy during shoulder arthroscopy in the beach chair and lateral decubitus positions. Anesth Analg. 2010;111:496–505
73. 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
74. Gill AS, Rajneesh KF, Owen CM, Yeh J, Hsu M, Binder DK. Early optical detection of cerebral edema in vivo. J Neurosurg. 2011;114:470–7
75. Dunham CM, Ransom KJ, Flowers LL, Siegal JD, Kohli CM. Cerebral hypoxia in severely brain-injured patients is associated with admission Glasgow Coma Scale score, computed tomographic severity, cerebral perfusion pressure, and survival. J Trauma. 2004;56:482–9
76. Leal-Noval SR, Cayuela A, Arellano-Orden V, Marín-Caballos A, Padilla V, Ferrándiz-Millón C, Corcia Y, García-Alfaro C, Amaya-Villar R, Murillo-Cabezas F. Invasive and noninvasive assessment of cerebral oxygenation in patients with severe traumatic brain injury. Intensive Care Med. 2010;36:1309–17
77. Bhatia R, Hampton T, Malde S, Kandala NB, Muammar M, Deasy N, Strong A. The application of near-infrared oximetry to cerebral monitoring during aneurysm embolization: a comparison with intraprocedural angiography. J Neurosurg Anesthesiol. 2007;19:97–104
78. Yokose N, Sakatani K, Murata Y, Awano T, Igarashi T, Nakamura S, Hoshino T, Katayama Y. Bedside monitoring of cerebral blood oxygenation and hemodynamics after aneurysmal subarachnoid hemorrhage by quantitative time-resolved near-infrared spectroscopy. World Neurosurg. 2010;73:508–13
79. Banaji M, Mallet A, Elwell CE, Nicholls P, Cooper CE. A model of brain circulation and metabolism: NIRS signal changes during physiological challenges. PLoS Comput Biol. 2008;4:e1000212
80. 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
81. Zweifel C, Castellani G, Czosnyka M, Carrera E, Brady KM, Kirkpatrick PJ, Pickard JD, Smielewski P. Continuous assessment of cerebral autoregulation with near-infrared spectroscopy in adults after subarachnoid hemorrhage. Stroke. 2010;41:1963–8
82. Li Z, Wang Y, Li Y, Wang 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
83. Fladby T, Bryhn G, Halvorsen O, Rosé I, Wahlund M, Wiig P, Wetterberg L. Olfactory response in the temporal cortex of the elderly measured with near-infrared spectroscopy: a preliminary feasibility study. J Cereb Blood Flow Metab. 2004;24:677–80
84. Zeller JB, Herrmann MJ, Ehlis AC, Polak T, Fallgatter AJ. Altered parietal brain oxygenation in Alzheimer’s disease as assessed with near-infrared spectroscopy. Am J Geriatr Psychiatry. 2010;18:433–41
85. Sokol DK, Markand ON, Daly EC, Luerssen TG, Malkoff MD. Near infrared spectroscopy (NIRS) distinguishes seizure types. Seizure. 2000;9:323–7
86. Watanabe E, Nagahori Y, Mayanagi Y. Focus diagnosis of epilepsy using near-infrared spectroscopy. Epilepsia. 2002;43 Suppl 9:50–5
87. Watson NF, Dodrill C, Farrell D, Holmes MD, Miller JW. Determination of language dominance with near-infrared spectroscopy: comparison with the intracarotid amobarbital procedure. Seizure. 2004;13:399–402
88. Obrig H, Steinbrink J. Non-invasive optical imaging of stroke. Philos Transact A Math Phys Eng Sci. 2011;369:4470–94
89. Swartling J, Bassi A, D’Andrea C, Pifferi A, Torricelli A, Cubeddu R. Dynamic time-resolved diffuse spectroscopy based on supercontinuum light pulses. Appl Opt. 2005;44:4684–92
90. Mesquita RC, Durduran T, Yu G, Buckley EM, Kim MN, Zhou C, Choe R, Sunar U, Yodh AG. Direct measurement of tissue blood flow and metabolism with diffuse optics. Philos Transact A Math Phys Eng Sci. 2011;369:4390–406