Neuromonitoring in the intensive care unit. I. Intracranial pressure and cerebral blood flow monitoring

Abstract

Monitoring the injured brain is an integral part of the management of severely brain injured patients in intensive care. Brain-specific monitoring techniques enable focused assessment of secondary insults to the brain and may help the intensivist in making appropriate interventions guided by the various monitoring techniques, thereby reducing secondary brain damage following acute brain injury.
This review explores methods of monitoring the injured brain in an intensive care unit, including measurement of intracranial pressure and analysis of its waveform, and techniques of cerebral blood flow assessment, including transcranial Doppler ultrasonography, laser Doppler and thermal diffusion flowmetry.
Various modalities are available to monitor the intracranial pressure and assess cerebral blood flow in the injured brain in intensive care unit. Knowledge of advantages and limitations of the different techniques can improve outcome of patients with acute brain injury.

Full text

Intensive Care Med (2007) 33:1263–1271
DOI 10.1007/s00134-007-0678-z
REVIEW
Anuj Bhatia
Arun Kumar Gupta
Neuromonitoring in the intensive care unit.
I. Intracranial pressure and cerebral b lood flow
monitoring
Received: 18 March 2007
Accepted: 22 March 2007
Published online: 24 May 2007
© Springer-Verlag 2007
Part II of this article is available at: http://
dx.doi.org/10.1007/s00134-007-0660-9.
A.Bhatia·A.K.Gupta(u)
Addenbrooke’s Hospital, Department of
Anaesthesia,
Hills Road, CB2 2QQ Cambridge, UK
e-mail: akg01@globalnet.co.uk
Tel.: +44-1223-217434
Fax: +44-1223-217223
A. K. Gupta
Addenbrooke’s Hospital, Neuroscience
Critical Care Unit,
Hills Road, CB2 2QQ Cambridge, UK
Abstract Background: Monitor-
ing the injured brain is an integral
part of the management of severely
brain injured patients in intensive
care. Brain-specific monitoring tech-
niques enable focused assessment of
secondary insults to the brain and
may help the intensivist in making
appropriate interventions guided by
the various monitoring techniques,
thereby reducing secondary brain
damage following acute brain injury.
Discussion: This review explores
methods of monitoring the injured
brain in an intensive care unit, in-
cluding measurement of intracranial
pressure and analysis of its wave-
form, and techniques of cerebral
blood flow assessment, including
transcranial Doppler ultrasonography,
laser Doppler and thermal diffusion
flowmetry. Conclusions: Various
modalities are available to monitor
the intracranial pressure and assess
cerebral b lood flow in the injured
brain in intensive care unit. Knowl-
edge of advantages and limitations of
the different techniques can improve
outcome of patients with acute brain
injury.
Keywords Traumatic brain injury ·
Intracranial pressure · Ultra-
sonography, Doppler, transcranial ·
Flowmetry, laser Doppler · Thermal
diffusion flowmetry
Abbreviations ABP: arterial blood
pressure · AMP: amplitude of
fundamental component of ICP
waveform · CBF: cerebral blood
flow · CBV: cerebral blood volume ·
CO
2
: carbon dioxide · CPP:
cerebral p erfusion pressure · CSF:
cerebrospinal fluid · CT:
computerised tomography · DID:
delayed ischaemic neurological
deficit · EDP: effective downstream
pressure · FV: flow velocity · HDI:
haemodynamic impairment · ICP:
intracranial pressure · ICU: intensive
care unit · LDF: laser Doppler
flowmetry · MAP: mean arterial
pressure · MCA: middle cerebral
artery · MRI: magnetic resonance
imaging · nCPP: non-invasively
determined cerebral perfusion
pressure · nICP: non-invasive ICP
measurement · PET: positron
emission tomography · PI: pulsatility
index · PRx: pressure reactivity
index · RI: resistance index · RAP:
correlation of amplitude and pressure
of ICP waveform · SAH:
subarachnoid h aemorrhage · TBI:
traumatic brain injury · TCD:
transcranial Doppler · TD: thermal
diffusion · THRT: transient
hyperaemic response test · US:
ultrasound · ZFP: zero flow pressure

1264
Introduction
Whilst there is little that neurointensive care can offer
to prevent the brain damage sustained from the primary
insult, the use of appropriate protocol-driven management
can, however, minimise the effects of secondary insults
on outcome [1]. Monitoring the injured brain is an in-
tegral part of the management of severely brain injured
patients in intensive care and can be classified into general
(systemic) and brain-specific methods. Although general
systemic monitoring (e.g. invasive arterial blood pressure,
end-tidal carbon dioxide, oxygen saturation of blood) is
of vital importance in detecting gross global changes in
physiology, brain-specific monitoring techniques enable
more focused assessment of secondary insults to the
brain and may help the intensivist in making appropriate
interventions guided by the various monitoring techniques.
An important adjunct to these techniques is the use
of a variety of imaging modalities, which, as a result of
significant advances over the last two decades, allow us to
assess the brain with respect to structural abnormalities,
blood flow and metabolism. The advantage of these
techniques [computerised tomography (CT), magnetic
resonance imaging (MRI), xenon-CT, positron emission
tomography, single photon emission computerised emis-
sion tomography, magnetic resonance spectroscopy] is that
they allow us to assess the interindividual heterogeneity
by providing detailed information of different regions of
the brain. However this information is not continuous and
at present cannot be obtained at the bedside.
This review focuses on brain-specific monitoring and
will include aspects of intracranial pressure measurement
and its interpretation, and methods to monitor cerebral
blood flow.
Intracranial pressure monitoring
Intracranial pressure (ICP) is the pressure within the cra-
nial vault relative to the ambient atmospheric pressure and
is now regarded as a core monitoring parameter in the in-
tensive care management of p atients with acute brain in-
jury. The ICP increases when compensatory mechanisms
which control ICP, such as changes in cerebrospinal fluid
(CSF) dynamics, cerebral blood flow (CBF) and cerebral
blood volume (CBV), are exhausted.
Whilst routine ICP monitoring is widely accepted as
a mandatory monitoring technique for management of
patients with severe head injury and is a guideline sug-
gested by the European Brain Injury Consortium [2], there
is some debate over its efficacy in improving outcome
from severe traumatic brain injury. A survey of Canadian
neurosurgeons revealed that only 20.4% of the respondents
had a high level of confidence in ICP monitoring [3], and
a survey of neurointensive care units in the UK showing
that only 75% of centres monitor ICP may reflect some
of the drawbacks of ICP monitoring [4]. A recently
published trial in survivors beyond 24 h following severe
brain injury that compared ICP-targeted intensive care
with management based on clinical observations and
CT findings reported no improvement in outcome with
ICP monitoring [5]. However, a review of neurocritical
care and outcome from traumatic brain injury (TBI) sug-
gested that ICP-/cerebral perfusion pressure (CPP)-guided
therapy may benefit patients with severe head injury,
including those presenting with raised ICP in the absence
of a mass lesion and also patients requiring complex
interventions [1].
Measurement of ICP
ICP can be measured at different sites in the brain–intra-
ventricular and intraparenchymal measurements are more
common, while extradural and subdural sensors are used
occasionally. Intraventricular catheters are still thought
of as the “gold standard” [6] as they allow direct meas-
urement by insertion of a catheter into one of the lateral
ventricles, which is connected to an external pressure
transducer [7]. The advantages of these systems are that
the clinician can check for zero drift and sensitivity of
the measurement system in vivo. Access to the ventric-
ular system also allows CSF drainage if the ICP rises.
However, this interferes with ICP monitoring, and only
one currently available catheter allows concomitant CSF
drainage and ICP monitoring (Rehau, Switzerland). The
drawbacks of such catheters include difficulty or failure
of insertion in patients with advanced brain swelling, as
the ventricles can be narrowed or effaced. An increase
in risk of infection after a period of time is another
potential problem, with reported rates of up to 10% [8]
with modern ventricular micro-transducers even though
these have excellent metrological properties [9]. There
are now commercially available ventricular catheters
with antibiotic-coated tips [Codman Bactiseal
®
external
ventricular drain catheter] that may reduce infection rates
but more studies are required before their use in clinical
practice can be supported.
The intraparenchymal systems may be inserted through
a support bolt or tunnelled subcutaneously from a bu rr
hole. These have a micro-miniature strain gauge pressure
sensor side-mounted at the tip (Codman) or a fibre-optic
catheter (Camino, Innerspace). Change of pressure results
in a change of resistance in the former and an alteration in
reflection of the light beam in the latter. Intraparenchymal
probes are a good alternative to ventricular catheters and
have a low infection rate [10], but in one study a signifi-
cant increase in colonisation at 5 days after insertion was
reported [11]. The main problem with these catheters is
a small drift of the zero line. Neither of these systems
allows pressure calibration to be performed in vivo.
After these systems are zeroed relative to a tmospheric
pressure during a pre-insertion calibration, their output

1265
is dependent on the zero drift of the sensor. Technical
complications such as kinking of the cable and dislocation
of the sensor have also been reported [12]. It should be
remembered that these sensors reflect a local pressure
value that may be misleading, as the ICP is not uniform
within the skull, e.g. supratentorial measurements may
not reflect infratentorial pressure. However, this is also
a problem with intraventricular catheters.
Subdural catheters are easily inserted following cran-
iotomy but measurements are unreliable because when
ICP is e levated, they are likely to underestimate the true
ICP. These are also liable to blockage. Extradural probes
have the advantage of avoiding penetration of the dura
but are even more unreliable as the relationship between
ICP and pressure in the extradural space is unclear. The
Gaeltec ICP/B solid-state miniature ICP transducers are
designed for use in the epidural space and are reusable,
and the zero reference can be checked in vivo. However,
measurement artefacts and decay in measurement quality
with repeated use have limited the acceptance of this
technology [13]. Despite these drawbacks, the subdural
and epidural catheters are associated with a lower risk
of infection, epilepsy and haemorrhage than ventricular
catheters [14, 15].
The Spiegelberg ICP monitoring system is a fluid-
filled catheter-transducer system that measures ICP using
a catheter that has an air-pouch balloon situated at the
tip. This device zeroes automatically in vivo and has
shown lower zero drift than standard catheter-tip ICP de-
vices [16]. This system can be used in epidural, subdural,
intraparenchymal and intraventricular sites. Despite its
obvious advantages, this system is still not used widely
and requires further evaluation.
ICP monitoring has several important applications,
some of which are discussed below:
Determination of CPP
Management of acute brain injury is largely CPP directed.
ICP is an important determinant of CPP [CPP = mean arte-
rial pressure (MAP)–ICP], which in turn affects CBF and
CBV. The optimal level of CPP is still under some debate,
with earlier studies recommending CPP > 70 mmHg [16]
although many other centres maintain CPP between 60
and 70 mmHg and one centre permits CPP as low as
50 mmHg [17].
ICP correlation with outcome
Experience from various centres with expertise in ICP
monitoring and research into TBI confirms that mean ICP
correlates with outcome with a threshold in the region of
25 mmHg. However no prospective study has been un-
dertaken (and is unlikely) to prove this and Brain Trauma
Foundation guidelines recommend ICP treatment should
be initiated at an upper threshold of 20–25 mmHg [18].
Waveform analysis
Analysis of ICP waveforms can be used to obtain infor-
mation about brain compliance. A computer program to
correlate mean ICP and AMP has been developed [19].
The ICP waveform consists of three components, which
overlap in the time domain but can be separated in the
frequency domain. The pulse waveform has several
harmonic components; the fundamental component has
a frequency equal to the heart rate. The amplitude of this
component (AMP) is very useful for the evaluation of
various indices. The linear correlation coefficient RAP
(R = symbol of correlation, A = amplitude, P = p ressure)
describes the relationship between pulse amplitude of
ICP and mean ICP value over short p eriods of time
(1–3 min). When RAP is positive, changes in AMP are in
the same direction as changes in mean ICP. When RAP
is negative, the change in AMP is reciprocal to those in
mean ICP value. Lack of synchronisation between fast
changes in amplitude and mean ICP is depicted by a RAP
of 0. A potential application of this model is to predict
outcome after severe head injury. This is possible because
of the nature of relationship between mean ICP and
AMP—as the ICP increases, the linear correlation with
AMP b ecomes distorted by an upper breakpoint which is
associated with a decrease in RAP coefficient from +1 to
negative values. A similar relationship between AMP and
ICP is seen in patients with severe brain injury. A plot o f
RAP against ICP shows similar results—RAP is positively
correlated to ICP in patients with good outcomes, whereas
the correlation decreases above ICP values of 20 mmHg
and becomes negative above 50 mmHg in patients who die
(Fig. 1) [20]. In the latter group of patients, the decrease
in RAP from +1 to 0 or negative precedes the final d e-
crease in ICP pulse amplitude and is a sign of impending
brainstem herniation.
Cerebrovascular pressure reactivity and derived indices
Cerebrovascular pressure reactivity, which is the change
in basal tone of smooth muscle in cerebral arterial walls
in response to changes in transmural p ressure, can be es-
timated from the ICP waveform by deriving the pressure
reactivity index (PRx). PRx has been used to determine
time responses to intracranial hypertension or changes in
mean arterial blood pressure in brain-injured patients [21].
PRx is calculated as a linear correlation coefficient be-
tween averaged arterial blood pressure and ICP from a time
window of 3–4 min. Good cerebrovascular reactivity is
associated with negative PRx and a poor reactivity with
a positive PRx value (Fig. 2). The PRx may be analysed as

1266
Fig. 1 Pulse amplitude of intracranial pressure (ICP)
(AMP, fundamental harmonic component) increases
with mean ICP until critical threshold is reached,
above which it starts to decrease. The correlation
coefficient between AMP and ICP (RAP) marks this
threshold by decreasing from positive to negative
values [20]
Fig. 2 PRx index, calculated as linear correlation coefficient between averaged ABP and ICP. Good cerebrovascular reactivity is associated
with negative PRx (a) and poor reactivity with positive PRx (b) [21]

1267
a time-dependent variable, responding to dynamic events
such as ICP plateau waves or incidents of arterial hypo-
and hypertension. The validity of PRx for monitoring and
quantifying cerebral vasomotor reactiv ity has been stud-
ied in patients with brain injury. A close link was found
between cerebral blood flow and intracranial pressure in
head-injured patients. This suggested that increases in arte-
rial blood pressure and cerebral p erfusion pressure may be
useful for reducing intracranial pressure in selected brain-
injured patients, i.e. those with intact cerebral vasomotor
reactivity [22].
There are drawbacks of monitoring ICP. It requires an
invasive procedure and personnel to monitor as well as to
react to changes. ICP monitoring is frequently performed
by non-neurointensivists. A survey of intensive care units
(ICU) in non-neurosurgical centres in the UK revealed that
though more than half of all such ICUs were admitting pa-
tients with severe TBI, only 9% used ICP monitoring as
a routine [23]. This has significant implications, especially
as there is a lack of class I evidence about the efficacy of
ICP monitoring in reducing morbidity and mortality.
It is possible that uptake of ICP monitoring may
increase if a non-invasive technique can be used. There
is currently keen interest in development of non-invasive
methods of ICP monitoring which include the use of
transcranial Doppler ultrasonography [24]. An example of
this is a procedure for continuously simulated ICP derived
from simultaneously recorded curves of ABP and flow
velocity (FV) in the middle cerebral artery (MCA) that has
been validated in patients with TBI [25]. This approach in-
volves a dynamic systems analysis technique and enables
modelling of physiological systems in which the inner
structure is too complex to b e d escribed mathematically.
The validity of this model was confirmed during infusion
studies in patients with hydrocephalus [26]. Use of such
non-invasive ICP (nICP) measurement techniques may
make ICP monitoring accessible to a wider range of ICUs.
Assessment of blood flow
Transcranial Doppler ultrasonography
Transcranial Doppler ultrasonography (TCD) is an ex-
tremely useful method for non-invasively monitoring
cerebral haemodynamics, by measuring red cell FV in real
time u sing the Doppler shift principle. Ultrasound (US)
waves are generated using a 2-MHz pulsed Doppler instru-
ment. In order to penetrate the skull, the same transducer
is used both for transmitting and receiving wave energy at
regular intervals. The moving blood acts as a reflector, first
receiving the transmitted wave from the transducer and
then reflecting it back. FV is calculated using the formula
for Doppler shift. Changes in FV correlate with changes
in CBF only if the angle of insonation and the diameter
of the insonated vessel remain constant [27]. Data are
generally derived from the MCA as it is easy to insonate,
carries a large proportion o f supratentorial blood and its
location allows easy fixation of the probe (to keep the
angle of insonation constant) for prolonged monitoring.
The transtemporal window through the thin bone above
the zygomatic arch is commonly used to insonate the
proximal segment (M1) of the MCA. In each patient, the
same insonation window should be used throughout the
entire study period.
As the volume of blood flowing through a vessel de-
pends on the velocity of the moving cells and the diameter
of the vessel concerned, then for a given blood flow, the
velocity will increase with decreases in vessel diameter.
Figure 3 shows a diagrammatic representation of a typi-
cal TCD waveform from the MCA. Mean FV (FV
mean
)is
the weighted mean velocity that takes into account the dif-
ferent velocities of the formed elements in the blood ves-
sel insonated and is normally around 55 ± 12 cm s
–1
.This
represents the most physiological correlate with the actual
CBF. The time-mean FV refers to the mean value of FV
max
and is determined from the area under the spectral curve.
The shape of the envelope (maximal shift) of the
Doppler spectrum from peak systolic flow to end-
diastolic flow with each cardiac cycle is known as the
waveform pulsatility. The FV waveform is determined
by the ABP waveform, the viscoelastic properties of
the cerebral vascular bed and blood rheology. In the
absence of vessel stenosis or vasospasm, changes in
ABP or blood rheology, the pulsatility reflects distal
cerebrova scular resistance. This resistance is usually
quantified by the pulsatility index (PI or Gosling index):
PI = (FV
systolic
–FV
diastolic
)/FV
mean
. Normal PI ranges
from 0.6 to 1.1 with no significant side-to-side or cerebral
interarterial differences and shows better correlation with
Fig. 3 Method of determining systolic (V
s
), diastolic (V
d
)andtime-
av eraged mean flow velocity (V
mean
) from the spectral outline. FV
is flow velocity and PI is pulsatility index [PI =(V
s
–V
d
)/V
mean
].
(Reproduced with permission from Greenwich Medical Media. In:
Gupta AK, Summors A, Notes in Neuroanaesthesia and Critical
Care, 2001)

1268
CPP than ICP. Another index that can be used to quantify
vessel resistance is the resistance index (RI or Pourcelot
index): RI = (FV
systolic
–FV
diastolic
)/FV
systolic
.
Applications of TCD
There are many advantages of using TCD. It is non-
invasive, relatively inexpensive and provides real-time
information with high temporal resolution. Some of the
clinical applications of TCD include the following:
Assessment of cerebral autoregulation and vasoreactivity
TCD is used in assessment o f cerebral autoregulation
and vasoreactivity to carbon dioxide (CO
2
)aslossof
these mechanisms in patients with brain injury may
indicate a poor prognosis. Autoregulation can be tested by
response of the TCD trace to vasopressor infusion (static
autoregulation) or thigh tourniquet deflation (dynamic
autoregulation). Autoregulation may also be tested at
the bedside using the transient hyperaemic response test
(THRT) [28], which assesses the hyperaemic response
in the TCD waveform following 5–9 s of digital carotid
artery compression. Lam et al. found that following an
aneurysmal subarachnoid haemorrhage, patients with
an initial impairment of the response to THRT were
more likely to develop delayed ischaemic neurological
deficits (DIDs) than patients with a normal response [29].
In a study using induced oscillations in the ABP (by
controlling ventilation) and calculating the phase shifts
between FV (measured using TCD) and ABP, cerebral
autoregulation was found to be impaired preceding the
onset of clinical vasospasm [30].
Detection of vasospasm following subarachnoid
haemorrhage
TCD is often used in the clinical setting to determine
presence of vasospasm. The primary effect of a decrease
in vessel lumen diameter is an increase in flow resistance,
and this results in an increase in FV. An MCA FV
mean
above 120 cm s
–1
is regarded as being significant [31] and
may indicate either hyperaemia or vasospasm. Although
it is generally regarded that vasospasm is likely if the
ratio of MCA FV to extracranial ICA FV (Lindegaard
ratio) is greater than 3 [32] and hyperaemia is present
if MCA FV > 120 cm s
–1
with a Lindegaard ratio less
than 3, the distinction is not well defined, especially
when commonly used TCD indices for diagnosis of
vasospasm are compared with cerebral perfusion findings
using PET. PET scans of patients following subarach-
noid haemorrhage (SAH) who developed DIDs showed
a wide range of cerebral perfusion disturbances, with
TCD indices failing to indicate these changes [33]. Thus
detection of vasospasm on TCD may not be associated
with delayed cerebral ischaemia and vice versa. Care
must therefore be taken when interpreting TCD data, and
these should be matched with clinical findings, and other
investigations such as xenon-CT flow measurements may
help to improve prediction o f vasospasm and hence avoid
repeated angiography [34]. However, when compared with
angiography for the MCA, TCD has been shown to give
high levels of specificity and positive predictive value for
vasospasm. Ratsep et al. found that vasospasm detected
by TCD is associated with haemodynamic impairment
(HDI, defined as blood flow velocity values consistent
with vasospasm in conjunction with impaired THRT);
thus, detection of HDI could identify patients at risk for
ischaemic complications [35].
Role of TCD in management of traumatic brain injury
Following traumatic brain injury, TCD monitoring can be
used to observe changes in FV, waveform pulsatility and
for testing cerebral vascular reserve. The autoregulatory
“threshold” or “breakpoint” (the CPP at which autoregula-
tion fails), which provides a target CPP value for treatment,
can also be determined by continuously recording the FV
from the MCA. At very low levels of CPP, as in brain
death, the microcirculation collapses. The net blood flow
diminishes, and the TCD pattern either shows low flow or
reve rsed flow during diastole.
Non-invasive determination of CPP
There is currently much interest in the use of TCD for non-
invasive determination of CPP (nCPP). This involves esti-
mation of CPP from parameters derived from MCA FV
and the ABP [24]. Schmidt et al. found that absolute dif-
ference between real CPP (i.e. MAP–ICP) and nCPP (i.e.
determined using TCD) was less than 13 mmHg in the ma-
jority of measurements for a range of CPPs between 60
and 100 mmHg. Such a difference may have significant im-
plications in patients with raised ICP (and possibly lower
CPP). The absolute value of side-to-side (i.e. interhemi-
spheric) difference in nCPP was significantly greater when
CT evidence of brain swelling was present and was also
correlated with mean ICP [36]. This technique needs to
be evaluated in further randomised trials focusing on its
accuracy, cost-effectiveness and validity before it can be
recommended for routine u se.
Measurement of zero flow pressure
A recent development is the use of TCD to measure zero
flow pressure (ZFP), i.e. the pressure at which CBF ceases,

1269
which gives an estimate of the effective downstream pres-
sure (EDP) of the cerebral circulation (Fig. 4). EDP, rather
than ICP, is believed to determine the effective CPP in the
absence of intracranial hypertension. FV in the MCA is
measured by TCD. EDP is derived from the ZFP as extra-
polated by regression analysis of instantaneous ABP/MCA
FV relationships. Buhre et al. reported that extrapolation of
ZFP enables detection of elevated ICP in patients with se-
vere head injury [37]. Thees et al. studied the correlation
between critical closing pressure determined using TCD
and ICP measured invasively. They found that using ICP
to determine CPP might overestimate the effective CPP,
i.e. the difference between MAP and CCP [38]. Further
evaluation is needed before this non-invasive technique of
measuring CPP can be accepted as a standard.
Confirmation of brain death
TCD has been suggested as a highly specific and sensi-
tive test for confirmation of brain death. It can be a useful
method to confirm brain death in patients in whom tradi-
tional brain death criteria cannot be used because of pos-
sible residual effects of sedative drugs [39]. The transtem-
poral approach is commonly used but the transorbital ap-
proach has also been successfully employed [40].
The limitations of using TCD are that it requires a cer-
tain degree of technical expertise, is operator dependent,
and the skull thickness, which varies with age, gender and
Fig. 4 Direct assessment of cerebral zero flow pressure. In a patient
with severe intracranial hypertension, no flow was observed in the
middle cerebral artery during diastole. The epidurally measured ICP
was 48 mmHg. (Reproduced with permission from [37])
race, may cause problems with transmission of ultrasound.
In fact, 10% of normal subjects cannot be assessed due
to lack of an adequate temporal window. The incidence of
failure can be reduced by increasing the power and perhaps
by use of 1 MHz probes [41]. In addition, TCD monitoring
focuses on the major cerebral a rteries but flow character-
istics in the cerebral microcirculation may be quite differ-
ent to those in the major arteries. Despite these limitations,
TCD holds promise of further applications for real-time
indirect assessment of CBF, non-invasive ICP and CPP.
Laser Doppler flowmetry
Laser Doppler flowmetry (LDF) allows continuous real-
time measurements of local microcirculatory blood flow
(red cell flux) with good dynamic resolution. Doppler
shift of reflected monochromatic laser light induced by
movement of red blood cells within the microcirculation
is measured. The magnitude and frequency distribution
of the wavelength changes are directly related to the
number and velocity of red blood cells but unrelated to the
direction of their movement. A 0.5–1 mm diameter fibre-
optic laser probe is placed in contact with or within brain
tissue and conducts scattered light back to a photodetector
within the flowmeter sensor. The signal is processed to
give a continuous voltage fluctuation versus time which is
linearly proportional to the real blood flow [42]. The probe
can be positioned in proximity to an area of intracranial
injury to monitor pathologic variations of microvascular
blood flow.
LDF is considered an excellent technique for instanta-
neous, continuous and real-time measurements of regional
CBF and for assessment of relative regional CBF changes.
LDF has a quick response to fluctuations in tissue per-
fusion and is relatively inexpensive. The relationship be-
tween LDF and CPP h as been found to change with time,
and this can indicate an improvement or deterioration in
autoregulation [43]. The main drawbacks of this technique
are that it is not a quantitative measure of CBF and mea-
sures CBF in a small brain volume (1–2 mm
3
). It is inva-
sive and prone to artefacts produced by patient movement
or probe displacement, which limits its clinical applicabil-
ity. However, it is a useful measure of local microcircula-
tory changes in combination with other monitoring tech-
niques and has been used to assess autoregulation, CO
2
reactivity and responses to therapeutic interventions [44]
and to detect ischaemic insults [45, 46].
Thermal diffusion flowmetry
Thermal conductivity of cerebral cortical tissues varies
proportionally with CBF, and measurement of thermal
diffusion (TD) at the cortical surface can be used for
CBF determination [47]. A monitor that measures TD

1270
flowmetry consists of two small metal plates, which are
thermistors, one of which is heated. Insertion of a TD
probe on the surface of the brain at a cortical region of
interest allows CBF to be calculated from the temperature
difference between the plates. An intraparenchymal TD
probe has also been evaluated and the results are encour-
aging [48]. Although the changes in CBF are relative, the
probes may be calibrated against absolute methods such
as xenon-133 or xenon-CT measurement of CBF to give
absolute values that assess blood flow changes in a small
volume of brain (± 20–30 mm
3
). Placement of the sensor
over large surface vessels should be avoided. Similar
sensors have been used to guide therapy for patients with
severe brain injury and intracerebral haematomas [49, 50].
Animal studies by Vajkoczy et al. revealed that TD
microprobes provide continuous real-time assessment of
intraparenchymal regional blood flow that was compar-
able with measurement by xenon-enhanced CT [51]. TD
flowmetry was characterised by more favourable diagnos-
tic reliability and was reported to be more sensitive than
TCD ultrasonography in assessing patients with reversible
vasospasm following intra-arterial injection of papaverine
in patients with SAH [52].
TD flowmetry has the potential for bedside monitor-
ing of cerebral perfusion at the tissue level, but it is inva-
sive and more clinical trials are needed to validate its use.
The intraparenchymal probes have excellent temporal res-
olution and it is possible that in the future a large part of
a single vascular territory may be monitored with single or
multiple probes.
This review has explored methods of assessing and
measuring intracranial pressure and cerebral blood flow
in the injured brain in the intensive care unit. Appropriate
use and knowledge of benefits and limitations of these
techniques can improve the outcome of patients with acute
brain injury.
References
1. Patel HC, Menon DK, Tebbs S,
Hawker R, Hutchinson PJ, Kirk-
patrick PJ (2002) Specialist neurocriti-
cal care and outcome from head injury.
Intensive Care Med 28:547–553
2. Maas AI, Dearden M, Serv adei F, Stoc-
chetti N, Unterberg A (2000) Current
recommendations for neurotrauma.
Curr Opin Crit Care 6:281–292
3. Sahjpaul R, Girotti M (2000) Intracra-
nial pressure monitoring in severe
traumatic brain injury—results of
a Canadian survey. Can J Neurol Sci
27:143–147
4. Wilkins IA, Menon DK, Matta BF
(2001) Management of comatose
head-injured patients: are we getting
any better? Anaesthesia 56:350–352
5. Cremer OL, van Dijk GW, van
Wensen E, Brekelmans GJ, Moons KG,
Leenen LP, Kalkman CJ (2005) Effect
of intracranial pressure monitoring and
targeted intensive care on functional
outcome after severe head injury. Crit
Care Med 33:2207–2213
6. The Brain Trauma Foundation. The
American Association of Neuro-
logical Surgeons. The Joint Section
on Neurotrauma and Critical Care
(2000) Recommendations for intracra-
nial pressure monitoring technology.
J Neurotrauma 17:497–506
7. Miller JD (1989) Measuring ICP in
patients: its value now and in the
future. In: Hoff JT, Betz AL (eds)
Intracranial pressure VII. Springer,
Berlin Heidelberg New York, pp 5–15
8. Bekar A, Goren S, Korfali E, Aksoy K,
Boyaci S (1998) Complications of
brain tissue pressure monitoring with
a fibreoptic device. Neurosurg Rev
21:254–259
9. Czosnyka M, Czosnyka Z, Pickard JD
(1996) Laboratory testing of three
intracranial pressure microtransduc-
ers: technical report. Neurosurgery
38:219–224
10. Mack WJ, King RG, Ducruet AF, Kre-
iter K, Mocco J, Maghoub A, Mayer S,
Connolly ES Jr (2003) Intracranial
pressure following aneurysmal sub-
arachnoid hemorrhage: monitoring
practices and outcome data. Neurosurg
Focus 14:1–5
11. Ghajar J (1995) Intracranial pressure
monitoring techniques. New Horiz
3:395–399
12. Stendel R, Heidenreich J, Schilling A,
Akhavan-Sigari R, Kurth R, Picht T,
Pietila T, Suess O, Kern C, Meisel J,
Brock M (2003) Clinical evaluation
of a new intracranial pressure moni-
toring device. Acta Neurochir (Wien)
145:185–193
13. Morgalla MH, Cuno M, Mettenleiter H,
Will BE, Krasznai L, Skalej M,
Bitzer M, Grote EH (1997) ICP mon-
itoring with a re-usable transducer:
Experimental and clinical evaluation of
the Gaeltec ICT/b pressure probe. Acta
Neurochir (Wien) 139:569–573
14. Gaab MR, Heissler HE, Ehrhardt K
(1989) Physical characteristics of
various methods for measuring ICP.
In: Hoff JT, Betz AL (eds) Intracra-
nial pressure VII. Springer, Berlin
Heidelberg New York, pp 16–21
15. Raabe A, Totzauer R, Meyer O,
Stockel R, Hohrein D, Schoeche J
(1998) Reliability of extradural pres-
sure measurement in clinical practice:
behaviour of three modern sensors
during simultaneous ipsilateral in-
traventricular or intraparenchymal
pressure measurement. Neurosurgery
43:306–311
16. Czosnyka M, Czosnyka Z, Pickard JD
(1997) Laboratory testing of the
Spiegelberg brain pressure monitor:
a technical report. J Neurol Neurosurg
Psychiatry 63:732–735
17. Grände PO, Asgiersson B, Nord-
ström CH (1997) Physiological
principles for volume regulation of
a tissue enclosed in a rigid shell
with application to the injured brain.
J Trauma 42(Suppl):S23–S31
18. Eker C, Asgeirsson B, Grande PO,
Schalen W, Nordstrom CH (1998)
Improved outcome after severe head
injury with a new therapy based on
principles for brain volume regulation
and preserved microcirculation. Crit
Care Med 26:1881–1886
19. Czosnyka M, Guazzo E, Whitehouse H,
Smielewski P, Czosnyka Z, Kirk-
patrick P, Piechnik S, Pickard JD (1996)
Significance of intracranial pressure
waveform analysis after head injury.
Acta Neurochir (Wien) 138:531–541
20. Balestreri M, Czosnyka M, Steiner LA,
Schmidt E, S mielewski P, Matta B,
Pickard JD (2004) Intracranial hyper-
tension: what additional information
can be derived from ICP waveform after
head injury? Acta Neurochir (Wien)
146:131–141

1271
21. Czosnyka M, Smielewski P, Kirk-
patrick P, Laing RJ, Menon D,
Pickard JD (1997) Continuous as-
sessment of the cerebral vasomotor
reactivity in head injury. Neurosurgery
41:11–19
22. Lang E W, Lagopoulos J, Grif-
fith J, Yip K, Yam A, Mudaliar Y,
Mehdorn HM, Dorsch NW (2003)
Cerebral vasomotor reactivity testing
in head injury: the link between pres-
sure and flow. J Neurolog Neurosurg
Psychiatry 74:1053–1059
23. McK eating EG, Andrews PJ, Tocher JI,
Menon DK (1998) The intensive care
of severe head injury: a survey of
non-neurosurgical centres in the United
Kingdom. Br J Neurosurg 12:7–14
24. Czosnyka M, Matta BF, Smielewski P,
Kirkpatrick PJ, Pickard JD (1998) Cere-
bral perfusion pressure in head injured
patients: a non invasive assessment
using transcranial Doppler ultrasonog-
raphy. J Neurosurg 88:802–808
25. Schmidt B, Czosnyka M, Schwarze JJ,
Sander D, Gerstner W, Lumenta CB,
Pickard JD, Klingelhofer J (1999)
Cerebral vasodilatation causing acute
intracranial hypertension: a method for
non-invasive assessment. J Cereb Blood
Flow Metab 19:990–996
26. Schmidt B, Czosnyka M, Schwarze JJ,
Sander D, Gerstner W, Lumenta CB,
Klingelhofer J (2000) Evaluation of
a method for non-invasive intracranial
pressure assessment during infusion
studies in patients with hydrocephalus.
J Neurosurg 92:793–800
27. Dahl A, Russell D, Nyberg-Hansen R,
Rootwelt K (1992) A comparison of
regional cerebral blood flow and middle
cerebral artery blood flow velocities:
simultaneous measurements in healthy
subjects. J Cereb Blood Flow Metab
12:1049–1054
28. Smielewski P, Czosnyka M, Iyer V,
Piechneik S, Whitehouse H, Pickard JD
(1995) Computerised transient hyper-
aemic response test—a method for
assessing autoregulation. Ultrasound
Med Biol 21:599–611
29. Lam JM, Smielweski P, Czosnyka M,
Pickard JD, Kirkpatrick PJ (2000)
Predicting delayed ischaemic deficits
after aneurysmal subarachnoid haem-
orrhage using a transient hyperaemic
response test of cerebral autoregulation.
Neurosurgery 47:819–825
30. Lang EW, Diehl RR, Mehdorn HM
(2001) Cerebral autoregulation testing
after aneurysmal subarachnoid haemor -
rhage: the phase relationship between
arterial blood pressure and cerebral
blood flow velocity. Crit Care Med
29:158–163
31. Jarus-Dziedzic K, Bogucki J, Zub W
(2001) The influence of ruptured cere-
bral aneurysm localization on the blood
flow velocity ev aluated by transcranial
Doppler ultrasonography. Neurol Res
23:23–28
32. Lindegaard KF, Nornes H, Bakke SJ,
Sorteberg W, Nakstad P (1988) Cere-
bral vasospasm after subarachnoid
haemorrhage investigated by means of
transcranial Doppler ultrasound. Acta
Neurochir (Wein) 42:81–84
33. Minhas PS, Menon DK, Smielewski P,
Czosnyka M, Kirkpatrick PJ, Clark JC,
Pickard JD (2003) Positron emission
tomographic cerebral perfusion dis-
turbances and transcranial Doppler
findings among patients with neuro-
logical deterioration after subarach-
noid haemorrhage. Neurosurgery
52:1017–1024
34. Horn P, Vajkoczy P, Bauhuf C,
Munch E, Poeckler-Schoeniger C,
Schmiedek P (2001) Quantitative
regional cerebral blood flo w tech-
niques improve non-invasive detection
of cerebrovascular vasospasm after
aneurysmal subarachnoid haemorrhage.
Cerebrovasc Dis 12:197–202
35. Ratsep T, Asser T (2001) Cerebral
haemodynamic impairment after
aneurysmal subarachnoid hemorrhage
as evaluated using transcranial Doppler
ultrasonography: relationship to de-
layed cerebral ischemia and clinical
outcome. J Neurosurg 95:393–401
36. Schmidt EA, Czosnyka M, Gooskens I,
Piechnik S K, Matta BF, Whitfield PC,
Pickard JD (2001) Preliminary e xperi-
ence of the estimation of the estimation
of cerebral perfusion pressure using
transcranial Doppler ultrasonography.
J Neurolog Neurosurg Psychiatry
70:198–204
37. Buhre W, Heinzel FR, Grund S,
Sonntag H, Weyland A (2003) Ex-
trapolation to zero-flow pressure in
cerebral arteries to estimate intracranial
pressure. Br J Anaesth 90:291–295
38. Thees C, Scholz M, Schaller C, Gass A,
Pavlidis C, Weyland A (2002) Rela-
tionship between intracranial pressure
and critical closing pressure in patients
with neurotrauma. Anesthesiology
96:595–599
39. Hadani M, Bruk B, Ram Z, Knoller N,
Spiegelmann R, Segal E (1999)
Application of transcranial Doppler ul-
trasonography for the diagnosis of brain
death. Intensive Care Med 25:822–828
40. Lampl Y, Gilad R, Eschel Y, Boaz M,
Rapoport A, Sadeh M (2002) Diagnos-
ing brain death using the transcranial
Doppler with a transorbital approach.
Arch Neurol 59:58–60
41. Klotzsch C, Popescu O, Berlit P (1998)
A new 1 MHz probe for transcranial
Doppler sonography in patients with
inadequate temporal bone windows.
Ultrasound Med Biol 24:101–103
42. Bolognese P, Miller JI, Heger IM,
Milhorat TH (1993) Laser Doppler
flowmetry in neurosurgery. J Neurosurg
Anesthesiol 5:151–158
43. Lam JM, Hsiang JN, Poon WS
(1997) Monitoring of autoregulation
using laser Doppler flowmetry in
patients with head injury. J Neurosurg
86:438–445
44. Kirkpatrick PJ, Smielweski P, Piech-
nik S, Pickard JD, Czosnyka M (1996)
Early effects of mannitol in patients
with head injuries assessed using
bedside multimodality monitoring.
Neurosurgery 39:714–720
45. Kirkpatrick PJ, Smielweski P, Czos-
nyka M, Pickard JD (1994) Continuous
monitoring of cortical perfusion by
laser Doppler flowmetry in ventilated
patients with head injury. J Neurolog
Neurosurg Psychiatry 57:1382–1388
46. Le Bihan D, T urner R (1992) The
capillary network: a link between IVIM
and classical perfusion. Magn Reson
Med 27:171–178
47. Carter LP (1991) Surface monitoring
of cerebral cortical blood flow. Cere-
brovasc Brain Metab Rev 3:246–261
48. Vajkoczy P, Roth H, Horn P, Lucke T,
Thome C, Hubner U, Martin GT,
Zappletal C, Klar E, Schilling L,
Schmiedek P (2000) Continuous mon-
itoring of regional cerebral blood flow:
experimental and clinical validation of
a normal thermal diffusion microprobe.
J Neurosurg 93:265–274
49. Carter LP, Weinand ME, Oommen KJ
(1993) Cerebral blod flow (CBF)
monitoring in intensive care by thermal
diffusion. Acta Neurochir Suppl (Wien)
59:43–46
50. Choksey MS, Chambers IR, Jenkins A,
Mendelow AD, Sengupta RP (1993)
Cortical thermal clearance monitoring
in surgery for giant middle cerebral
artery aneurysm. Br J Neurosurg
7:673–676
51. Vajkoczy P, Horn P, Thome C, Munch E,
Schmiedek P (2003) Regional cerebral
blood flow monitoring in the diag-
nosis of delayed ischemia following
aneurysmal subarachnoid hemorrhage.
J Neurosurg 98:1227–1234
52. Vajkoczy P, Horn P, Bauhuf C,
Munch E, Hubner U, Thome C (2001)
Effect of intra-arterial papaverine on
regional cerebral blood flow in hemody-
namically relevant cerebral vasospasm.
Stroke 32:498–505