Cerebral oxygen monitoring with near infrared spectroscopy: clinical application to neonates

Role of Cerebral Oximetry in Extracorporeal Membrane Oxygenation

Cerebral oximetry, which is based on near-infrared spectroscopy (NIRS) technology, is an optical technique that allows for noninvasive and continuous monitoring of brain oxygenation by determining cerebral tissue blood oxygen saturation. Many research and observational studies were performed with neonates using various types of NIRS/cerebral oximetry monitors. However, no food and drug administration (FDA) approved-cerebral oximeter is available for neonates. Successful validation of cerebral oximetry for the FDA has been done in human adult volunteer studies under protocols in which jugular bulb and arterial blood samples were obtained under different levels of fractional inspired oxygen levels.

Cerebral Oxygenation Monitoring: A Strategy to Detect IVH and PVL

Premature infants are at risk for intraventricular hemorrhage (IVH) and periventricular leukomalacia (PVL) theorized to be a result from fluctuations in cerebral blood flow. Monitoring cerebral oxygenation offers a method to observe changes in cerebral blood flow that may be beneficial for detecting and preventing IVH and PVL. This article explains the potential for cerebral oxygenation monitoring in detecting IVH and PVL using cerebral oximetry, reviews current knowledge known about cerebral oxygenation, and describes current challenges for cerebral oxygenation to be the next neuroprotective vital sign.

Cerebral oxygenation during hypoxia and resuscitation by using near-infrared spectroscopy in newborn piglets

BACKGROUND

Hypoxic events and cardiac arrest may cause brain damage in critical infants. This study investigated cerebral tissue oxygenation and oxygen extraction in a piglet model of hypoxic events, cardiac arrest and effects of resuscitation.

METHODS

For the hypoxia experiment, anesthetized newborn piglets were randomized to a hypoxia group (n = 8) with decreasing ventilatory rate to 0, and a control group (n = 8) with no hypoxic conditions. Regional cerebral tissue oxygen saturation (rScO2, detected by near-infrared spectroscopy) and oxygen saturation were recorded every 5 minutes for 100 minutes. Fractional cerebral tissue oxygen extraction (FTOE) was calculated as (arterial oxygen saturation [SaO2] - rScO2)/SaO2. For the resuscitation experiment, animals were grouped as hypoxia-no CPR (n = 4), control-no CPR (n = 4), and control-CPR (n = 4) after cardiac arrest. Standard cardiopulmonary resuscitation (CPR) was performed on the control-CPR group and observed for 30 minutes.

RESULTS

Immediate and significant changes in rScO2, and gradual changes in FTOE were observed during the hypoxia experiment. In the resuscitation experiment, no significant differences in rScO2 were found between groups. However, the highest FTOE was observed in the control-CPR group.

CONCLUSION

Noninvasive monitoring of rScO2 and evaluating FTOE changes during hypoxia and resuscitation may help clinicians evaluate brain tissue oxygenation and viability.

Validation of a noninvasive neonatal optical cerebral oximeter in veno-venous ECMO patients with a cephalad catheter

INTRODUCTION

Cerebral Oximetry is an optical technique that allows for noninvasive and continuous monitoring of brain oxygenation by determining tissue oxygen saturation (SctO2). In conjunction with pulse oximetry, cerebral oximetry offers a promising method to estimate cerebral venous oxygen saturation (SvO2).

OBJECTIVE

The aim of this study was to validate the cerebral oximetry measurements with the cerebral oxygen saturation measured from blood drawn in neonates on veno-venous ECMO with existing cephalad catheter with a prototype neonatal cerebral oximeter developed by CAS Medical Systems (Branford, CT, USA).

STUDY DESIGN

After obtaining informed consent, neonates undergoing VV-ECMO with cephalad catheterization were monitored by the CAS cerebral oximeter. Cephalad blood samples were periodically obtained to validate the monitor's accuracy.

RESULTS

Seventeen neonates were studied with 1718 h of cerebral oximetry data collected. Compared to the reference values, the bias+/-precision for cerebral oximetry SctO2 was 0.4+/-5.1% and derived SvO2 was 0.6+/-7.3%.

CONCLUSION

We recommend the use of this noninvasive method as an alternative to blood draws for cerebral venous saturation measurements in neonates requiring extracorporeal life support.

In vivo near-infrared spectroscopy

Interrogation of tissue with light offers the potential for noninvasive chemical measurement, and penetration with near-infrared wavelengths (750-1000 nm) is greater than with visible light. Specific absorption by clinically relevant compounds such as oxy- and deoxyhemoglobin and the intracellular respiratory enzyme cytochrome oxidase enable in vivo measurement of these to be performed safely and conveniently. This is the basis of in vivo near-infrared spectroscopy (ivNIRS). Multiple scattering of the interrogating beam by tissues leads to an optical path that is considerably longer than the simple physical pathlength and this complicates the analysis. Modeling of photon propagation through tissues with, for example, finite element and Monte Carlo methods, is assisting in improving the ivNIRS methodology. Instrumentation has advanced from simple continuous wave approaches, through time-resolved methods based on either time-domain or frequency-domain approaches, to spatially resolved measurement based on diffuse reflectance. Initial clinical applications were for monitoring the brain in the neonate and fetus and muscle in adults. Currently, use in adults and children for neurological assessments are of growing interest.

Brain perfusion monitoring with frequency-domain and continuous-wave near-infrared spectroscopy: a cross-correlation study in newborn piglets

The newborn piglet brain model was used to correlate continuous-wave (CW) and frequency-domain (FD) near-infrared spectroscopy. Six ventilated and instrumented newborn piglets were subjected to a series of manipulations in blood oxygenation with the effects on brain perfusion known to be associated with brain hypoxia-ischaemia. An excellent agreement between the CW and FD was demonstrated. This agreement improved when the scattering properties (determined by the FD device) were employed to calculate the differential pathlength factor, an important step in CW data processing.

Impact of hypoxemia on the performance of cerebral oximeter in volunteer subjects

Adverse neurological events during hypoxic episodes in high-risk patients or in patients not thought to be at risk while undergoing procedures increase morbidity and mortality. The ability to reliably monitor cerebral oxygenation could serve as an indicator for the need of therapeutic intervention and it's overall effect. This study was designed to verify the reliability of the only commercially available continuous noninvasive monitor, the INVOS 3100 (Somanetics Corp., Troy, MI), in subjects with varying levels of hypoxemia. Six adult volunteer subjects were enrolled. After placement of electrocardiogram (EKG), noninvasive blood pressure (NIBP), pulse oximeter (SpO2), cerebral oximeter (rSO2), a 20 g radial artery catheter, and a 4 F oximetric jugular bulb catheter, the subjects were given hypoxic mixtures to breathe to varying levels of desaturation. Arterial and mixed venous blood was drawn for blood-gas analysis at each level of O2 saturation. The cerebral hemoglobin saturation value from the cerebral oximeter was compared to the combined brain saturation using the formula: estimated field saturation between the light source and the detector (fSO2) = 0.25 x the arterial oxygen saturation (SaO2) + 0.75 x the jugular bulb venous oxygen saturation (SjvO2), (fSO2 = 0.25 SaO2 + 0.75 SjvO2). Statistical analysis demonstrated a correlation of 0.67 between rSO2 and fSO2 and a bias of -3.1% with a precision of 12.1%. Minimal bias of 0.38% and precision of 6.22% were calculated for transitional error. We concluded from the study that rSO2 may serve as a reliable indicator of changes in brain oxygenation induced by hypoxemia.

Calibration of near-infrared frequency-domain tissue spectroscopy for absolute absorption coefficient quantitation in neonatal head-simulating phantoms

Frequency-domain tissue spectroscopy is a method to measure the absolute absorption coefficient of bulk tissues, assuming that a representative model can be found to recover the optical properties from measurements. While reliable methods exist to calculate absorption coefficients from source-detector measurements less than a few centimeters apart along a flat tissue volume, it is less obvious what methods can be used for transmittance through the larger tissue volumes typically associated with neonatal cerebral monitoring. In this study we compare the use of multiple distance frequency-domain measurements processed with (i) a modified Beer-Lambert law method, (ii) an analytic infinite-medium diffusion theory expression, and (iii) a numerical finite element solution of the diffusion equation, with the goal of recovering the absolute absorption coefficient of the medium. Based upon our observations, the modified Beer-Lambert method provides accurate absolute changes in the absorption coefficient, while analytic infinite-medium diffusion theory solutions or finite element-based numerical solutions can be used to calculate the absolute absorption coefficient, assuming that the data can be measured at multiple source-detector distances. We recommend that the infinite-medium multi-distance method or the finite element method be used across large tissue regions for calculation of the absolute absorption coefficient using frequency-domain near-infrared measurements at multiple positions along the head.

Role of frequency domain optical spectroscopy in the detection of neonatal brain hemorrhage--a newborn piglet study

OBJECTIVE

Inability of continuous wave (CW) optical spectroscopy to measure changes in scattering, and the use of an arbitrary rather than an actual baseline, makes the CW method highly susceptible to errors that can lead to a false-positive or false-negative diagnosis. Our objective was to assess whether, and to what extent, the use of quantitative frequency domain spectroscopy would improve our ability to detect and monitor the development of brain hemorrhage.

METHODS

A dual-channel frequency-domain tissue spectrometer (Model 96208, ISS, Inc., Champaign, IL) was used to monitor the development of experimental subcortical and periventricular-intraventricular hemorrhage (IVH) in 10 newborn piglets (blood injection model). The multidistance approach was employed to calculate the absorption and reduced scattering coefficients and hemoglobin changes from the ac, dc, and phase values acquired at four different source-detector distances and at 752 nm and 830 nm.

RESULTS

There were significant absorption and scattering changes in the subcortical hematoma (n = 5) and the IVH groups (n = 5). The smallest detectable amount of blood in the brain was 0.04 ml. Changes associated with subcortical hematoma were several times higher than those associated with IVH, and correlated better with the estimated cross-sectional area of the hematoma than with the volume of the injected blood. As opposed to IVH, there was a significant absorption difference between the injured (subcortical hematoma) and normal side of the brain, probably because in case of IVH a significant volume of the injected blood had accumulated/spread beyond the reach of the probe.

CONCLUSION

Clearly, frequency-domain spectroscopy cannot increase our ability to quantify the volume (size) or the oxygenation of the injected blood, especially in the case of IVH. However, the ability to quantify the baseline tissue absorption and scattering would significantly improve diagnostic performance, and may allow for early identification and treatment of neonatal brain hemorrhage.

Neonatal cerebral oxygen regulation after hypothermic cardiopulmonary bypass and circulatory arrest

OBJECTIVE

Despite technical advances, neurologic sequelae continue to occur in neonates after heart surgery using deep hypothermic cardiopulmonary bypass (dhCPB) and circulatory arrest (DHCA). This study sought to determine the cerebral microcirculatory responses to hypoxia, hypotension, hypocapnia, and hypercapnia after dhCPB and DHCA.

DESIGN

Prospective laboratory animal trial.

SETTING

Research laboratory.

SUBJECTS

Twenty-eight newborn pigs.

INTERVENTIONS

Piglets were divided into control, dhCPB, and DHCA groups. The control group received surgery. The dhCPB group received surgery and deep hypothermic CPB for 40 mins. The DHCA group received surgery, deep hypothermic CPB for 40 mins, and circulatory arrest for 60 mins. Two hours after the intervention, cerebral microcirculatory responses were examined.

MEASUREMENTS AND MAIN RESULTS

Cerebral microcirculatory responses consisted of changes in cerebral oxygen saturation (Sco2) and pial arteriolar diameter measured by near- infrared spectroscopy and intravital microscopy, respectively. All groups experienced similar decreases in Sco2 and increases in pial arteriolar diameter in response to moderate and severe hypoxia (Pao2, 35 and 25 torr, respectively) and moderate and severe hypotension (mean pressure, 30 and 20 mm Hg, respectively). Sco2 and pial arteriolar diameter decreased to hypocapnia (Paco2, 25 torr) similarly in all groups. To hypercapnia (Paco2, 70 torr), Sco2 increased in the control group, did not change in the dhCPB group, and decreased in the DHCA group. Pial arteriolar diameter to hypercapnia increased in the control and the dhCPB groups but did not change in the DHCA group.

CONCLUSIONS

Cerebral vascular and oxygenation responses to hypoxia, hypocapnia, and hypotension were preserved after dhCPB and 1 hr of DHCA. By comparison, cerebral vascular and oxygenation responses to hypercapnia were not; both vascular and oxygenation responses were altered after DHCA, but only the oxygenation response was altered after dhCPB. These data suggest a selective disturbance in the microcirculation and/or parenchymal oxygen metabolism after DHCA and dhCPB.

The use of near infrared spectroscopy (NIRS) in children after traumatic brain injury: a preliminary report

Children commonly develop diffuse cerebral swelling after traumatic brain injury (TBI) which is believed due to a secondary response to the injury. Near infrared spectroscopy (NIRS), a continuous, direct, and noninvasive monitor of cerebral oxygenation and cerebral blood volume (CBV), could be helpful in understanding these secondary responses. The aims of our study were to determine whether NIRS used in children with severe TBI will provide insight into the pathophysiology of injury. Ten children (1 mo to 15 years old) with severe TBI (admission GCS < or = 7) were continuously monitored by NIRS by placing optodes over the frontalparietal region. Relative values of oxyhemoglobin (HbO2), deoxyhemoglobin (Hb), and total hemoglobin (THb) were obtained and compared to intracranial pressure (ICP), mean arterial pressure (MAP), electroencephalography (EEG), and arterial PCO2 (PaCO2). Episodes of intracranial hypertension (ICP > 20 Torr [T]), changes in ICP > 10 T, changes in PaCO2 > or = 8 T, and changes in MAP > 20 T frequently resulted in changes in HbO2, Hb, and THb. Hyperventilation with decreased PaCO2 always resulted in cerebral oxygen desaturation irregardless of ICP. Often, high ICP correlated with increased THb and HbO2 indicating increased CBV and cerebrovascular dilatation. In two children, posttraumatic seizures were preceded by an unexplained rapid cerebral hyperoxygenation several hours prior to the onset of the clinical seizures. NIRS reliably detects changes in cerebral hemodynamics in children and may be used to further understand the etiology of the diffuse cerebral swelling seen in children after severe TBI.

Near infrared spectroscopy (NIRS) in patients with severe brain injury and elevated intracranial pressure. A pilot study

Near infrared spectroscopy (NIRS) was used to asses changes in regional cerebral oxygen saturation (rSO2) in 8 head injured patients with an intracranial pressure (ICP) higher or lower than 25 mmHg (n = 4 for each group). NIRS values in the high ICP group (> 25 mmHg) were significantly lower than in the low ICP group (< 25 mmHg). In contrast, arterial pO2, pCO2, peripheral oxygen saturation and transcranial Doppler sonography (TCD) values were similar in both groups. To further investigate changes in rSO2 to changes in peripheral oxygen saturation and arterial pO2, patients of both groups underwent an artificial hyperoxygenation (50% O2) period of 3 minutes. Both groups revealed similar values in peripheral oxygen saturation, arterial pO2, and TCD velocities at the end of the hyperoxygenation period. However, rSO2 values in patients with an ICP > 25 mmHg were significant lower than in patients with an ICP < 25 mmHg after the hyperoxygenation period. In addition, patients with an ICP < 25 mmHg revealed a significant increase in rSO2 values at the end of the hyperoxygenation period, not detectable in patients with an ICP > 25 mmHg. Our results suggest that NIRS may be an additional diagnostic tool in the non-invasive evaluation of impaired cerebral microcirculation in patients with increased intracranial pressure.

Does the redox state of cytochrome aa3 reflect brain energy level during hypoxia? Simultaneous measurements by near infrared spectrophotometry and 31P nuclear magnetic resonance spectroscopy

We studied cerebral oxygen metabolism during hypoxia to demonstrate whether the redox state of cytochrome aa3 (cyt.aa3), as measured by near infrared spectrophotometry (NIRS), reflects brain energy level. Rats (n = 6) subjected to hypoxia were simultaneously monitored by NIRS and 31P nuclear magnetic resonance spectroscopy (NMRS). Brain function was evaluated using the electroencephalogram (EEG). After a reduction of the fraction of inspired oxygen FIO2 from 0.21 to 0.15, we observed a significant increase in reduced cyt.aa3 (from 16.5% +/- 2.1% to 41.2% +/- 2.8%; P < 0.01), without significant changes in phosphocreatine (PCr) and beta-adenosine triphosphate (beta-ATP) levels. The PCr decreased significantly at a FIO2 of 0.10 (53.8% +/- 6.4% as compared with 97.7% +/- 10.9% at a FIO2 of 0.21; P < 0.05), and reached a minimum at a FIO2 of 0.04. beta-ATP did not change significantly at a FIO2 of 0.10 or 0.08. With a FIO2 of less than 0.08, cyt.aa3 was almost totally reduced. EEG activity slowed at a FIO2 of 0.08 and became isoelectric at 0.04. Significant correlations were found between the levels of cyt.aa3 and PCr (P < 0.001, r = 0.83) as well as between cyt.aa3 and beta-ATP (P < 0.001, r = 0.73) by using the overall values at FIO2 levels from 0 to 1.0. However, no significant correlations were observed among these variables when the FIO2 was less than 0.10. These findings suggest that the increase in reduced cyt.aa3 reflects brain energy depletion; however, the redox state of cyt.aa3 will not indicate brain energy depletion during extreme hypoxia because cyt.aa3 is reduced totally during hypoxia insufficient to deplete intracellular ATP.

Near-infrared spectroscopy: theory and applications

In conclusion, NIRS appears to offer both a new monitoring modality and new information about cerebral oxygenation. Technical problems in the application of this technology persist, most notably determination of pathlength and the volume of tissue interrogated. Those familiar with the history of pulse oximetry will recall that although Millikan developed an ear oximeter in 1947, it was not until Aoyagi combined recognition of the pulse signal with spectroscopy in the 1970s that oximetry was transformed into a clinically applicable monitor. In much the same way, NIRS may find the same tremendous usefulness as a noninvasive monitor of cerebral oxygen utilization, pending resolution of the remaining technical problems.

Near infrared spectroscopy in adults. Does the Invos 3100 really measure intracerebral oxygenation?

Does the Invos near infrared spectroscopy monitor detect intracranial oxygenation, or reflect oxygenation of extracerebral tissues? The change in Invos regional cerebral oxygenation was studied during a doubling of middle cerebral artery flow with hypercapnia. There was no significant change in regional cerebral oxygenation, which suggests that the Invos does not reflect intracerebral oxygenation, possibly because of its narrow optode separation. The use of near infrared spectroscopy in adults may require a more powerful machine.

Near-infrared monitoring of the cerebral circulation

Near-infrared spectroscopy is a noninvasive bedside technique for monitoring hemoglobin saturation (HbO2%) in brain vasculature. The method linearly relates the optical signal detected from the surface of the head to HbO2%. To do so, the method relies on constant transcranial optical pathlength and light scattering as well as minimal interference by tissues overlying the brain. This study examined these premises. Optical signals from a dual-wavelength, near-infrared spectrometer were correlated with sagittal sinus HbO2% in 7 anesthetized piglets subjected to 7 different physiological conditions: normoxia, moderate and severe hypoxia, hyperoxia, hypocapnia, hypercapnic hyperoxia, and hypotension. These conditions were induced by varying the inspired O2 concentration (7-100%), ventilatory rate (5-35 breaths/min), and blood pressure (phlebotomy 20 ml/kg) to force HbO2% over a wide range (5-93%). To evaluate interference by tissues overlying the brain, correlations were repeated after the scalp and skull were rendered ischemic. Transcranial optical pathlength was measured by phase-modulated spectroscopy. Linear relationships between optical signals and sagittal sinus HbO2% were found with correlation coefficients ranging from -0.89 to -0.99 (p < 0.05) among animals; however, slope and intercept had coefficients of variability of approximately 15 and 333%, respectively. Almost identical linear expressions were observed whether scalp and skull were ischemic or perfused. Transcranial optical pathlength was constant in each animal, but ranged from 10 to 18 cm among animals. The data indicate that the assumptions underlying near infrared spectroscopy are reasonably accurate in a given animal, but that the constants for transcranial optical pathlength and light scattering are not the same in all animals.(ABSTRACT TRUNCATED AT 250 WORDS)

Near-infrared spectroscopic localization of intracranial hematomas

Near-infrared spectroscopy (NIRS) of the cerebral hemispheres, applied transcranially through the intact scalp and skull, was evaluated for its ability to detect the presence of an intracranial hematoma in 46 head-injured patients. In 40 patients intracranial hematomas (22 subdural, 10 epidural, eight intracerebral) were identified on computerized tomography (CT); in all 40 cases, NIRS demonstrated greater absorption of light at 760 nm on the side of the hematoma. The mean difference in optical density (OD) between the hemisphere with the hematoma and the normal hemisphere was 0.99 +/- 0.30 for epidural hematomas, 0.87 +/- 0.31 for subdural hematomas, but only 0.41 +/- 0.11 for intracerebral hematomas. In 36 patients, the asymmetry in OD resolved after surgical evacuation of the hematoma or with spontaneous resorption of the hematoma. Four patients who developed postoperative or delayed hematomas exhibited persistence of the asymmetry in OD. Six patients had only diffuse injuries and exhibited only minor differences in OD between the hemispheres, similar to 10 patients in the control group with no head injury. It appears that NIRS is useful in the initial examination of the head-injured patient, as an adjunct to CT, and in following patients postoperatively in the intensive care unit.