Transcranial doppler, EEG and SEP monitoring

Abstract

The role of neuromonitoring in the prevention of cerebral damage associated with cardiosurgical interventions has not yet been clearly elucidated. Reliable randomised studies from evidence-based medicine showing a clear reduction of risk do not exist. Numerous studies and reviews however, have confirmed that non-invasive procedures for monitoring neuronal or neurophysiological changes before, during and after interventions within the heart or the major thoracic vessels are available and provide early indications of damage.Technological modalities and clinical indications for non invasive cerebral monitoring were evaluated:Electroencephalography (EEG) with processed EEG, bispectral index (BIS) and the evoked potential for use with spinal cord functionNear infrared spectroscopy (NIRS) for assessment of cerebral perfusion and oxygenationTranscranial Doppler sonography (TCDS) for assessment of cerebral circulation and perfusionMultimodality monitoring as a combination of EEG, NIRS and TCDS.

Full text

224 H. Gehring, L. Meyer zu Westrup, S. Boye, A. Opp, U. Hofmann
Background
The role of neuromonitoring in the prevention of cere-
bral damage associated with cardiosurgical interven-
tions has not yet been clearly elucidated. Reliable ran-
domised studies from evidence-based medicine show-
ing a clear reduction of risk do not exist. Numerous
studies and reviews however, have confirmed that
non-invasive procedures for monitoring neuronal or
neurophysiological changes before, during and after
interventions within the heart or the major thoracic
vessels are available and provide early indications of
damage [1 – 10].
The incidence of cerebral damage with cardiac or
major thoracic vessel operations in adults and children
has not improved despite technological developments
in monitoring, surgery and perfusion (tab. 1 - 3). It
should also be stressed that up until now it has not
been possible to create a uniform nomenclature for
damage to or functional loss of neuronal structures, so
that a reliable form of ensuring comparability still re-
mains an enigma (tab. 4). As regards pathogenesis, nu-
merous covariables exist that may concern the condi-
tion of the patient, the surgical procedure, the perfu-
sion technique being used or indeed the mode of anes-
thesiological management (tab. 5).
Major problems that can lead to reversible, time/in-
tensity-dependent and even irreversible damage to
neural structures include:
– Ischaemia (embolisation, perfusion, hypoxia,
haemoglobin, oxygen saturation)
– Metabolism (hypoglycaemia, acidosis)
– Electrolyte displacement (hyponatremia)
Here, inadequate perfusion or defective oxygena-
tion can be detected, differentiated and treated early on
as potential factors leading to ischemia (tab. 6).
Transcranial doppler, EEG and SEP monitoring
H. Gehring1, L. Meyer zu Westrup1, S. Boye1, A. Opp2, U. Hofmann3
1Department of Anesthesiology, University Hospital of Schleswig-Holstein, Campus Luebeck, Luebeck, Ger-
many; 2Institute of Biomedical Engineering, University of Luebeck, Luebeck, Germany; 3Institute of Signal Pro-
cessing, University of Luebeck, Luebeck, Germany
Applied Cardiopulmonary Pathophysiology 13: 26-00, 2009
Keywords:neuromonitoring, cerebral damage, cardiac surgery
Abstract
The role of neuromonitoring in the prevention of cerebral damage associated with cardiosurgical interventions has
not yet been clearly elucidated. Reliable randomised studies from evidence-based medicine showing a clear re-
duction of risk do not exist. Numerous studies and reviews however, have confirmed that non-invasive procedures
for monitoring neuronal or neurophysiological changes before, during and after interventions within the heart or
the major thoracic vessels are available and provide early indications of damage.
Technological modalities and clinical indications for non invasive cerebral monitoring were evaluated:
– Electroencephalography (EEG) with processed EEG, bispectral index (BIS) and the evoked potential for use
with spinal cord function
– Near infrared spectroscopy (NIRS) for assessment of cerebral perfusion and oxygenation
– Transcranial Doppler sonography (TCDS) for assessment of cerebral circulation and perfusion
– Multimodality monitoring as a combination of EEG, NIRS and TCDS.

Transcranial doppler, EEG and SEP monitoring 225
Table 1: Cognitive decline after surgery with cardiopulmonary bypass in percent.
Observation period Reference
1 week 1 - 6 months 5 years
Newman 1995 73 37 [11]
Newman 2001 53 36/24 42 [12]
Van Dijk 2007 50.4 on-pump [13]
Van Dijk 2007 50.4 off-pump [13]
Table 2: Incidence of stroke in a 5-year postoperative period.
Reference
Van Dijk 2007 1.4 off-pump [13]
Van Dijk 2007 3.6 on-pump [13]
Table 3: Differentiated retrospective analysis of the incidence of stroke and delirium with respect to interventions [14].
2003 ∑
2 or 3 fold mitral aortic CABG CABG CABG
n = 16184 valves valve valve + valve on pump off-pump
% 4.6 9.7 8.8 4.8 7.4 3.8 1.9 stroke
% 8.4 11.2 7.9 2.3 delirium
Table 4: Current nomenclature for damage (left) or functional loss
(middle) of neuronal structures with the main resulting residuals
(right column).
Neurological injury Neurocognitive decline Delirium
Encephalopathy Neurocognitive dysfunction Seizures
Deterioration Stroke
Intellectual function Stupor
Memory deficite Coma
Table 5: Covariables with effects on neurophysiological deficiencies
Interventions Co-Morbidity Surgical procedures
Coronary artery bypass graft surgery Extra-, intracranial stenosis Hypothermia
Open chamber surgery Arrhythmia Normothermia
1, 2 or 3 valves Thrombogenic aortic arch Management arterial blood pressure
Congenital defects Age Off- vs on-pump
Aortic surgery at different levels Unstable angina Circulatory arrest
Endovascular procedures (stents, valves) Diabetes Separate perfusion

226 H. Gehring, L. Meyer zu Westrup, S. Boye, A. Opp, U. Hofmann
A range of monitoring procedures (tab. 7 and 8)
now allows the detection and assessment of this poten-
tial damage early on. Individual procedures have their
strengths and weaknesses; so that the effectiveness of
a therapeutic measure (and also the detection of a re-
versible course of damage) can only be described in
full if the procedures are combined with one another
(multimodality monitoring).
Transcranial Doppler Sonography (TCDS)
Using a 2 MHz pulsed-wave transducer, TCDS [3, 4 -
8, 16 - 23] can be performed through the temporal win-
dow for insonation of selected sections of intracranial
arterial vessels (tab. 9). The method measures direc-
tion and the velocity profile of erythrocytes (fig. 1).
The absolute values depend on the insonation angle.
The signal provides information about blood when the
Table 6: The virtuous circle of neurological damage.
Pathway
Ischemia
Metabolism
Electrolyte shift
Causes
embolisation
hypoglycemia
hyponatremia
perfusion
acidosis
hemoglobin
lactate
oxygen saturation
hypoxia
Diagnosis
TEE
Arterial blood gas analysis
Effect
Edema
ICP
Table 7: Characteristics of non invasive monitoring procedures
Ideal monitor Characters TCDS EEG pEEG BIS SSEP tcMEN NIRS
+++ noninvasive +++ ++ ++ +++ ++ (+) +++
+++ continuous ++ ++ +++ ++ + + +++
+++ objective ++ (-) (+) - - - +
+++ rapid + + + + (+) + +
+++ both hemispheres ++ + ++ - +++
+++ sensor handling - - (+) + (+) (+) ++
From ideal to difficult: +++; ++; +; (+); (-); -; = not intended
TCDS – Transcranial Doppler Sonography; EEG – Electroencephalography; pEEG – Processed Electroencephalography; BIS – Bispectral
Index Monitor; SSEP – Somatosensoric Evoked Potentials; tcMEN – Transcranial Motor Evoked Potentials; NIRS – Near Infrared Spec-
troscopy
Table 8: Synergistic effects of multimodality monitoring.
Ischemia
TCDS
EEG
NIRS
Perfusion
Blood flow
+++
(+)
+
Embolisation
+++
-
-
Oxygenation
Hemoglobin
-
+
++
Oxygen saturation
+++
Neuronal function
Reversibility vs. irreversibility
-
+++
-
Bihemispherial
+
++
+++
Multimodality Monitoring
Interpretation and decision
++
+
+++
Diagnosis and therapeutic intervention
+++
+++
+++

Transcranial doppler, EEG and SEP monitoring 227
insonation angle is held constant. With holding sys-
tems, one sensor each for the right and the left hemi-
spheres can be used continuously during an interven-
tion [21]. Systems are also available that are designed
for use during anaesthetic management (mask ventila-
tion, endotracheal intubation, and central venous ac-
cess) that provide comfort for the conscious patient
(fig. 2 and 3) [21]. The system failure rate lies between
3 [22] and 21 % [17], and primarily depends on the in-
vestigator. The signals provide clear data on vascular
resistance and the preservation of cerebral autoregula-
tion. The critical closing pressure (CCP), which is not
a constant value, can be assessed when reducing per-
fusion during a cardiopulmonary bypass. CCP is the
most relevant parameter when assessing regional cere-
bral perfusion.
Continuously bilateral TCDS recording detects
solid and gaseous emboli, whereby high intensity tran-
sient signals (HITS) denote severe events (fig. 4) [16].
If severe emboli events are detected, neuroprotective
management should be initiated [20]. An erroneous
aortal cannulation with a side-differential perfusion or
inadequate venous drainage can also be detected im-
mediately so that immediate corrective adjustments
can be introduced before any irreversible damage due
to deficient cerebral blood flow would otherwise set in
within minutes.
One limitation of TCDS is that insonation can on-
ly be performed on two vessel sections of interest, and
not on the complete circulating system. However, the
possibility to insonate vascular sections of the Circu-
lus Willis (fig. 5 and 6) does provide a benefit since
Table 9: Characteristics of the Transcranial Doppler sonography (TCDS)
Transcranial doppler sonography (TCDS)
Measure
cerebral
blood flow velocity
Limitations
failure - user dependent
10 - 21 %
relative blood flow changes
sensor fixation
closing pressure
technical equipment
vascular resistance
sample volume
vessel regions
Detect
emboli
gaseous/solid
Advantages
perfusion
emboli detection
Assess
autoregulation
control arterial canula
pediatric patients
Needs
windows
left and right temporal
2 sensors
constant insonation
Fig. 1: TCDS signal of the left mean cerebral
artery (MCA L) insonated at a depths of 56
mm (D56) with a perfusion index (PI) of 0.97
and a mean velocity (Vm) of 42 cm/sec.

228 H. Gehring, L. Meyer zu Westrup, S. Boye, A. Opp, U. Hofmann
Fig. 2: Multimodal cerebral monitoring with pEEG,
NIRS (one sensor) and bilateral TCDS before induction
of anesthesia.
Fig. 3: Degree of handling with the TCDS holding sys-
tem for use with anesthesia.
Fig. 4: Embolic signals of air (right) and high intensity transient signals (HITS, left).

Transcranial doppler, EEG and SEP monitoring 229
flow characteristics for central cerebral regions can
then be provided.
Cerebral autoregulation and perfusion pressure can
be assessed when arterial pressure is continuously
recorded in parallel (fig. 7). This allows the measure-
ment of the critical closing pressure in the selected
vessel regions (fig. 8).
Electroencephalography (EEG)
EEG reflects the electrical activity of the neuronal
structure of the brain [3, 6 - 8, 18 - 20]. Because of
physical restrictions on signal recording the method
only presents information on the cortex, with subcorti-
cal structures largely escaping any assessment (tab.
10). EEG is the most sensitive method for detecting
imbalances in cellular function, and can also provide
information about the reversibility of injury. It is also
the best procedure for detecting seizure activities of a
Fig. 5: Use of the left and right temporal window for in-
sonation of two vessel sections (MCA left and PCA with
P1 segment on the right).
Fig. 6: Corresponding sig-
nals revealed from insonation
of two vessel sections (MCA
left and PCA with P1 segment
on the right).
Fig. 7: Continuous recording of the TSDS
trends and the corresponding mean arteri-
al pressure for assessment of cerebral au-
toregulation and cerebral closing pres-
sure. The lines identified the closing pres-
sure when reducing blood flow (left) and
with restart (right). The parallel lines to
the x axis identified the corresponding
pressure.

230 H. Gehring, L. Meyer zu Westrup, S. Boye, A. Opp, U. Hofmann
(sub)clinical nature during anaesthesia. Depending on
the number of sensors fitted according to the standard
10/20 coordinate system (fig. 9), the investigator can
acquire either generalised or regionalised information.
Cerebral ischaemia induces neuronal dysfunction so
that frequencies are slowed or amplitudes are reduced
(tab. 11). The latency of return of EEG activity after
deep hypothermia cardiac arrest (DHCA) predicts the
subsequent neurological outcome [8].
The EEG signal depends substantially on thermo-
management, anaesthetic depth, imbalances in metab-
olism (hypoglycaemia), and autoregulation combined
with the arterial carbon dioxide partial pressure (pCO2
in mmHg). This dependence on anaesthetic depth has
led to the introduction of depth-of-anaesthesia moni-
toring, including the measurement of the bispectral in-
dex (BIS). Hypothermia may suppress cerebral activi-
ty due to burst suppression and might mask the depth
of anaesthesia or any severe cerebral ischaemia. Unex-
pected consciousness during a cardiopulmonary by-
pass with hypothermia is a serious anaesthetic compli-
cation [24].
EEG is less sensitive for detecting hypoxemia than
cerebral near infrared spectroscopy (NIRS), but more
sensitive in the detection of recovery. [8].
Electrode handling and signal interpretation are the
limitations of EEG. A comparison of baseline data with
signals obtained during intervention, and between
Fig. 8: Corresponding closing pressures of
the P1 segment (+) and the MCA (trian-
gle) with respect to cooling and rewarm-
ing.
Table 10: Characteristics of the Electroencephalography (EEG)
Electroencephalography (EEG)
Measure
electrical activity
Limitations
sensor dependent
amplitude
number
frequency
location
not subcortical structures
Detect
neuronal imbalance
interpretation
cortex
Advantages
neurophysiological imbalance
regional discrimination
reversibility before damage
Interference
anesthesia
pCO2
hypothermia
glucose
metabolism
electrolyte displacement

Transcranial doppler, EEG and SEP monitoring 231
those taken from the left and right hemispheres might
help to identify an ischaemia.
Processed EEG
Different forms of spectral analysis (processed EEG)
from EEG signals allow characterisation of the fre-
quency-power distribution for interpreting whether
signals are due to anaesthesia or ischemia, and for fa-
cilitating the provision of monitor information and sig-
nal display (tab. 12). Quantitative variables can be ex-
tracted from the spectra of the frequency and power
data. The spectral edge frequency (SEF 95) contains
95 % of the global power, whereas the median de-
scribes the 50 % edge. Quantitative power variables
are defined as the total power (representing the sum of
the total power of all frequencies), the relative power
for a given frequency band (α-, β-, δ-, and θ-band),
and the ratios to the total power.
Bispectral IndexTM
Unlike classical spectrum analysis, bispectral analysis
also quantifies the phase spectrum and reports a di-
mensionless BIS value of between 100 (awake) and 0
(no EEG-activity) [25, 26]. The BIS value allows an
assessment of sedation, sleep and the effects of anaes-
thesia. Changes can occur due to the influence of opi-
oids or hypothermia. Muscular artefacts disrupt the
crude EEG which is also displayed and filtered out
(tab. 13). Up until now mainly one sensor is used with
2 measurement electrodes and an indifferent electrode
on the forehead according to the equivalent positions
of the 10/20 scheme so that no hemisphere differences
are recorded. A sensor system assessing both hemi-
spheres is currently launched. Any unnoticed con-
sciousness during interventions with a heart lung ma-
chine or during hypothermia represents a clear danger;
it emphasises the importance of using the BIS monitor
to control the depth of anaesthesia rather than for mon-
itoring neuronal function.
Somatosensorically-evoked potentials (SSEP)
Somatosensorically-evoked potentials [3, 15, 27 – 31]
(tab. 14) are cortical and subcortical responses to the
stimulation of peripheral nerves (fig. 10). For the pur-
poses of signal production the data from individual
stimulations are added or averaged (> 100). Depend-
ing on the point of stimulation, early, middle and late
latency times are defined which are associated with
their corresponding conduction pathways. Positive
Fig. 9: International 10/20 sensor position system for
EEG
Table 11: Characteristic information to EEG bands
Comparison of EEG bands
Type
Frequency (Hz)
Differential diagnosis
Delta
up to 3
subcortical diffuse lesions
metabolic encephalopathy hydrocephalus
deep midline lesions.
Theta
4 - 7 Hz
focal subcortical lesions
metabolic encephalopathy
deep midline disorders
Alpha
8 - 12 Hz
coma
Beta
12 - 30 Hz
benzodiazepines
Gamma
26–100
Table 12: Characteristics of the processed EEG (pEEG)
Processed EEG (pEEG)
Measure
frequency-power distribution
quantitative variable
spectrale edge frequency
SEF 95
Assess
both hemispheres
frontal - occipital
Advantage
number of electrodes
4 channel
Limitation
objective interpretation

232 H. Gehring, L. Meyer zu Westrup, S. Boye, A. Opp, U. Hofmann
events are labelled with a P, and negative with an N,
while the subsequent number usually declares the la-
tency time of the event in msec. For the stimulus posi-
tion of the median nerve a typical negative wave N20
results, while the characteristic curve for the posterior
tibial nerve is labelled as P35. A reduction of ampli-
tude of 50 % and a prolongation of the latency time by
10 % indicates an ischaemia. SSEP-monitoring origi-
nates from intact sensory conduction pathways and is
compared with the baseline before the intervention
(10/50 rule). Wherever damage has occurred a quanti-
tative analysis is insufficient.
Transcranial motor evoked potentials (tcMEP)
In order to test the motor conduction pathways as well,
the stimulus responses can be derived as electrical po-
tentials from their corresponding muscles through
transcranial electrical or magnetic stimulation of the
cortex [27 – 30](tab. 15). While the extremely painful
electric stimulation can only be applied during deep
anaesthesia, transcranial magnetic stimulation can also
be applied during a postoperative follow-up. Stimulus
application and responses can be safely reached using
percutaneous needles under anaesthesia. Anaesthesia
procedures and neuromuscular blockade can signifi-
cantly alter the outcome and must be considered dur-
ing interpretation. Here as well the comparison with
baseline before the intervention is the key aspect of the
analysis.
Near-Infrared Spectroscopy (NIRS)
The procedure [5, 6, 8, 18, 19, 23, 33 - 35] (tab. 16) is
based on the subtraction of signals from 2 to 4 wave-
lengths in the range from 700 and 1000 nm and on the
Table 13: Characteristics of the Bispectral Index Monitoring (BIS)
Bispectral Index (BIS)
Measure
classical spectrum analysis
with phase spectrum
Reduce
number of electrodes
1 sensor
2 channel
electromyography
artefacts
Advantages
1 value between 100 and 0
Hypnosis and sedation
(avoid awareness)
resistent to artefacts
Limitations
Protection not evidence based
1 hemisphere
designed to measure anesthetic effect
Table 14: Characteristics of the Somatosensoric Evoked Potentials
(SSEP, SEP)
Somatosensoric Evoked Potentials (SSEP, SEP)
Measure
Assess
Advantage
Limitations
cortical and subcortical response to peripheral
nerve stimulation
200 averaged data sets
assessment of spinal cord injury
10/50 rule
sensoric pathway
pre/post changes
prognostic value to mortality
low sensitivity to
delayed neurological deficite
Fig. 10: Principles of Evoked Potentials (EP) for the
assessment of spinal cord injuries during aortic
aneurysm repair. TcMEP on the left side and SSEP on
the right side.

Transcranial doppler, EEG and SEP monitoring 233
ability to adequately penetrate tissues and skull bone
in the area applied. The subtraction is achieved in such
a way that in addition to an emitter there are also two
detectors lying one centimetre apart that record the
signal. Depending on the absorption by the tissue and
the path that the photons must cross, the proportion of
signal which penetrates the cerebral cortex in the mar-
ginal zones can be extracted. The so called „sample
volume“ of the sensor can vary very much between in-
dividuals. Absolute values can only be given for the re-
gional cerebrovenous oxygen saturation (rCVOS) in
percent, since this is a ratio obtained from the signals
of 2 wavelengths. The proportion of arterial and ve-
nous blood lies between 25/75 and 15/85 %, so that
differences from the measured jugular venous bulb
oxygen saturation (SjvO2 in %) can materialise if the
device is not calibrated to this figure. A clear diagno-
sis of a hypoxia, i.e. a defective oxygen partial pres-
sure in the tissue < 22 mmHg, can not be reliably made
from the oxygen binding curve. In addition the NIRS-
signal can also be contaminated by tissues and ex-
tracranial vessels.
Due to the easy handling and the ability to fix two
sensors on the forehead this non invasive monitoring
develops a career like the pulse oximetry: there is no
evidence based information about outcome benefit,
but it delivers clear physiological information about
tissue oxygenation at the sample volume. Effects of
perfusion restriction with associated ischemia and the
corresponding recovery with respect to differences be-
tween hemispheres can be monitored (fig. 11).
Multimodality monitoring
Simultaneous neurological monitoring – TCDS,
processed EEG and NIRS – may hold the greatest
promise for detecting and correcting neuronal dys-
function and later occurring neurological defects [2, 6,
19] (tab. 17). With combined use of TCDS to measure
blood flow in the middle cerebral artery (MCA) and
NIRS to measure rSO2 in the frontal lobe, and in both
hemispheres as well, it is possible to monitor up to 70
% of the blood flow distribution.
Monitoring perfusion with TCDS, oxygenation
with NIRS and the resulting neuronal imbalance with
pEEG can clearly identify the cause of injury and may
assist in choosing the appropriate corrective measures.
The highly sensitive EEG-alterations can help differ-
entiate between the effects of hypothermia and anaes-
thetic depth.
In patients with congenital heart defects, Austin et
al. [19] introduced a treatment algorithm. When the
abnormal monitoring values were treated, the inci-
dence of adverse outcome decreased to 7 %, similar to
the 6 % value obtained from patients without abnormal
values.
Multimodality neuromonitoring is also supported
by the fact that injury to the brain is a multifactorial
process.
Documentation
There is currently no standardized documentation
chart available for neurophysiologic monitoring and
Table 15: Characteristics of the Transcranial Motor Evoked Poten-
tials (tcMEP)
Transcranial motor evoked potentials (tcMEP)
Measure
Assess
Advantage
Limitations
response of M. tibialis anterior/gastrocnemicus
to central motor cortex stimulation
transcutaneuos magnetic stimulation
percutaneous stimulation with needles
painful
10/50 rule
descendence motoric pathway
postoperative magnetic stimulation
percutaneous needles
painful
Table 16: Characteristics of the Near-infrared Spectroscopy (NIRS)
Near-infrared Spectroscopy (NIRS)
Measure
Assess
Advantages
Limitations
tissue absorption
2 to 4 wavelength (LED or laser)
700 to 1000 nm
sensitive to oxygen saturation
hemoglobin
cytochrome a/a3
mixed venous/arterial saturation
both hemispheres
continuous signal
handling
small sample volume cortex
calibration procedure
relative changes
prognostic value

234 H. Gehring, L. Meyer zu Westrup, S. Boye, A. Opp, U. Hofmann
changes. With respect to the pre- and post interven-
tional periods and with the left and right central nerv-
ous structures we propose a brain chart (fig. 12).
Outlook
TCDS
TCDS is restricted by the necessary handling of the
probes – the use of a continuously applied sensor hold-
ing device or adding ultrasound gel. The application
can only be indicated with detection from emboli
where arterial cannulation of a thrombogenic aortal
arch has been carried out, or in patients with vascular
disease who require a higher cerebral perfusion pres-
sure or for whom the leading vessels are stenosed.
Amongst young patients with inborn cardiac defects
there is a much wider range of application due to the
reliable accessibility of the intracerebral vessels
through the temporal window, the information regard-
ing cerebral bloodflow and vascular resistance, as well
as the autoregulation. However, health insurances do
not cover its widespread usage in heart surgery in the
USA [lit].
Fig. 11: NIRS forehead sensors
during clamp of the left internal
carotic artery (ACI left). Graph
from M. Heringlake with permis-
sion).
Table 17: Characteristics of the Multimodality Monitoring
Mulimodality Monitoring
Measure
Assess
Advantages
Limitations
perfusion TCDS
oxygenation NIRS
neuronal imbalance EEG
up to 70 % blood flow
side-different effects to perfusion
incidence of embolic events
highly sensitive to oxygen saturation
hemoglobin
differential diagnosis
target treatment
reversibility
sensor group
skin space
additional monitors
experience with interpretation Fig. 13: Chart for perioperative documentation of
neurological monitoring and damages.

Transcranial doppler, EEG and SEP monitoring 235
EEG
Global EEG-analysis is not routinely applied during
cardiosurgical interventions. Because of the very sim-
ple application of pEEG and BIS, these procedures can
certainly be used for neuromonitoring, since it is also
possible to assess the depth of anaesthesia. Here as
well risk patients should receive special attention,
since information provided by the BIS procedure does
not currently allow any clear decisions to be made
about treatment. Despite all attempts to introduce and
establish dimensionless scales as single parameters,
use of the monitor depends largely on the anaes-
thetist’s personal experience.
NIRS
The simple handling, inexpensive electrodes and the
clear information provided about cerebral oxygenation
and perfusion (also where the two hemispheres are
compared) all go hand-in-hand to recommend this pro-
cedure as a standard for cardiosurgical intervention.
Adverse effects of neuromonitoring
The area on the forehead, the positioning and the space
required for the fitted sensors often present limitations
for the procedure, especially where congenital heart
surgery is being conducted in children.
Preparation of the skin and sensor pressure can
lead to irritation, although more often than not this is
self-restricted. Commercial holding devices are avail-
able for continuous bilateral TCDS recording. Poten-
tial effects on eyes and ears should be considered; the
patient should best be awake and able to seek help. All
sensors should be visible and checked periodically by
the anaesthesiologists.
Signal loss might be caused by perfusion or by a
displacement of the sensor on the skin. It therefore re-
quires some experience in TCDS to allow a fixed po-
sition of the sensors to be maintained.
Conclusion
Impairment of cerebral function is a multifactorial
process. The most important parameters for continu-
ous monitoring procedures are the comparisons:
1. to the baseline before intervention, and
2. between the two hemispheres.
The sensitivities of NIRS for ischemia, TCDS for
perfusion, EEG for neuronal function and EP for sen-
sory and motor neuronal function have all been ade-
quately proven, whereby multimodal application of
the procedures optimises the informative value regard-
ing cause, therapy and effect. The failure of TCDS and
EP alone stresses the importance of the simple and
standardised application of NIRS. The use of bilateral
and unilateral TCDS or SSEP/tcMEP is indicated for
special areas of application.
References
1. Roach GW. Adverse cerebral outcomes after coronary bypass
surgery. N Eng J Med 1996; 335:1857-63
2. Edmonds HL Jr. Multi-modality neurophysiologic monitoring
for cardiac surgery. Heart Surg Forum 2002; 5: 225-8
3. Sloan MA. Prevention of ischemic neurologic injury with intra-
operative monitoring of selected cardiovascular and cerebrovas-
cular procedures: Roles of electroencephalography, somatosen-
sory evoked potentials, transcranial doppler, and near-infrared
spectroscopy. Neurol Clin 2006; 24: 631-45
4. Murkin JM. Applied neuromonitoring and improving CNS out-
comes. Semin Cardiothorac Vasc Anesth 2005; 9: 139-142
5. Murkin JM. Neurocognitive outcome: The year in review. Curr
Opin Anaesthesiology 2005; 18: 57-62
6. Ghanayem NS. Monitoring the brain before, during, and after
cardiac surgery to improve long-term neurodevelopment out-
comes. Cardiol Young 2006; 16: S103-S109
7. Guarracino F Cerebral monitoring during cardiovascular sur-
gery. Curr Opin Anaesthesiology 2008; 21: 50-54
8. Hoffmann GM. Neurological monitoring on cardiopulmonary
bypass: What are we obligated to do? Ann Thorac Surg 2006;
81: S2373-80
9. Chan MT. Interventional neurophysiologic monitoring. Curr
Opin Anaesthesiology 2004; 17: 389-96
10. Edmonds HL Jr. Protective effect of neuromonitoring during
cardiac surgery. Ann N Y Acad Sci 2005; 1053: 12-9
11. Newman MF. Longitudinal assessment of neurocognitive func-
tion after coronary artery bypass grafting. N Engl J Med 2001;
344: 395-402
12. Newman MF. Central nervous system injury associated with car-
diac surgery. Lancet 2006; 19: 694-703
13. Van Dijk D. Cognitive and cardiac outcomes 5 years after off-
pump vs on-pump coronary artery bypass graft surgery. JAMA
2007; 701-8
14. Bucerius J. Predictors of delirium after cardiac surgery delirium:
effect of beating-heart (off-pump) surgery. J Thorac Cardiovasc
Surg 2004; 127: 57-64
15. Karkouti K. Low hematocrit during cardiopulmonary bypass is
associated with increased risk of perioperative stroke in cardiac
surgery. Ann Thorac Surg 2005; 80: 1381-7
16. Abu-Omar Y. Solid and gaseous cerebral microembolization
during off-pump, on-pump, and open cardiac surgery proce-
dures. J Thorac Cardiovasc Surg 2004; 127: 1759-65
17. Djaiani G. Epiaortic scanning modifies planned intraoperative
surgical management but not cerebral embolic load during coro-
nary artery bypass surgery. Anesth Analg 2008; 1006: 1611-8
18. Andropoulos DB. Neurological monitoring for congenital heart
surgery. Anesth Analg 2004; 99: 1365-75

236 H. Gehring, L. Meyer zu Westrup, S. Boye, A. Opp, U. Hofmann
19. Austin EH 3rd. Benefit of neurophysiologic monitoring for pe-
diatric cardiac surgery. J Thorac Cardiovasc Surg 1997; 114:
707-15
20. Yeh T Jr. Rapid recognition and treatment of cerebral air em-
bolism: the role of neuromonitoring. J Thorac Cardiovasc Surg
2003; 126: 589-91
21. Gehring H. A new probe holding device for continuous bilateral
measurements of blood flow velocity in basal brain vessels.
AINS 1997; 32: 355-9
22. Meyer zu Westrup L. Überwachung der zerebralen Integrität
während Eingriffen mit Herz-Lungen-Maschine und Hypother-
mie mit der bilateralen transkraniellen Dopplersonographie, der
zerebralen Nahe-Infrarot-Spektroskopie und der prozessierten
EEG-Analyse. Medical Thesis, University of Luebeck, 2000
23. Moritz S. Accuracy of cerebral monitoring in detecting cerebral
ischemia during carotid endarterectomy. Anesthesiology 2007;
107: 563-9
24. Avidan MS. Anesthesia Awareness and the Bispectral Index. N
Engl J Med 2008; 358: 1097-108
25. Rosow C. Bispectral index monitoring. Anesthesiol Clin North
America 2001; 19: 947-66
26. Vretzakis G. Influence of bispectral index monitoring on deci-
sion making during cardiac anesthesia. J Clin Anaesth 2005; 17:
509-16
27. Weigang E. Setup of neurophysiological monitoring with
tcMEP/SSEP during thoracoabdominal aneurysm repair. Thorac
Cardiov Surg 2005; 53: 28-32
28. Weigang E. Perioperative management to improve neurologic
outcome in thoracic or thoracoabdominal aortic stent grafting.
Ann Thorac Surg 2006; 82: 1679-87
29. Weigang E. Efficacy and frequency of cerebrospinal fluid
drainage in operative management of thoracoabdominal aortic
aneurysms. Thorac Cardiov Surg 2007; 55: 73-78
30. Sloan TB. Electrophysiologic monitoring during surgery to re-
pair the thoraco-abdominal aorta. J Clin Neurophysiol 2007; 24:
316-27
31. Achouh E. Role of somatosensory evoked potentials in predict-
ing outcome during thoracoabdominal aortic repair. Ann Thorac
Surg 2007; 84: 782-8
32. Yeh T Jr. Mixed venous oxygen saturation does not adequately
predict cerebral perfusion during pediatric cardiopulmonary by-
pass. J Thorac Cardiovasc Surg 2007; 122: 192-3
33. Steinrink J. Relevance of depth resolution for cerebral blood
flow monitoring by near-infrared spectroscopic tracking during
cardiopulmonary bypass. J Thorac Cardiovasc Surg 2006; 132:
1172-8
34. Edmonds HL Jr. Pro: all cardiac surgical patients should have in-
traoperative cerebral oxygenation monitoring. J Cardiothorac
Vasc Anesth 2006; 20: 445-9
35. Goldman S. Optimizing intraoperative cerebral oxygenation de-
livery using noninvasive cerebral oximetry decreases the inci-
dence of stroke for cardiac surgical patients. Heart Surg Forum
2004; 7: E376-E381
Address for corresponding: Hartmut Gehring, M.D., Ph.D., Dept.
Anesthesia, University of Lübeck, Ratzeburger Allee 160, 23538
Lübeck, Germany, E-Mail: gehring@uni-luebeck.de