European Journal of Anaesthesiology 2008; 25 (Suppl 42): 167–173

r 2008 Copyright European Society of Anaesthesiology

doi: 10.1017/S0265021507003341

Original Article

Transcranial Doppler ultrasonography in intensive care

F. A. Rasulo, E. De Peri, A. Lavinio

Institute of Anesthesiology and Intensive Care, Spedali Civili University Hospital of Brescia, Piazzale Spedali Civili,
Brescia, Italy

Summary
Transcranial Doppler is an innovative, flexible, accessible tool for the bedside monitoring of static and
dynamic cerebral flow and treatment response. Introduced by Rune Aaslid in 1982, it has become indis-
pensable in clinical practice. The main obstacle to ultrasound penetration of the skull is bone. Low frequencies,
1–2 MHz, reduce the attenuation of the ultrasound wave caused by bone. Transcranial Doppler also provides
the advantage of acoustic windows representing specific points of the skull where the bone is thin enough to
allow ultrasounds to penetrate. There are four acoustic windows: transtemporal, transorbital, suboccipital and
retromandibular. The identification of each intracranial vessel is based on the following elements: (a) velocity
and direction; (b) depth of signal capture; (c) possibility of following the vessel its whole length; (d) spatial
relationship with other vessels; and (e) response to homolateral and contralateral carotid compression. The
main fields of clinical application of transcranial Doppler are assessment of vasospasm, detection of stenosis of
the intracranial arteries, evaluation of cerebrovascular autoregulation, non-invasive estimation of intracranial
pressure, measure of effective downstream pressure and assessment of brain death. Mean flow velocity is
directly proportional to flow and inversely proportional to the section of the vessel. Any circumstance that
leads to a variation of one of these factors can thus affect mean velocity. The main pathological condition
affecting flow velocity is the vasospasm. Vasospasm is a frequent complication of subarachnoid haemorrhage, it
often remains clinically silent and the factors that make it symptomatic are largely unknown. Threshold
velocities above which vasospasm comes into place are well defined as regards the median cerebral artery, while
there is no consensus for the other vessels. Nevertheless, an increase in velocity alone is not sufficient to arrive
at a diagnosis of vasospasm; a condition of hyperaemia also presents with an increase in flow velocity. The
Lindegaard Index has therefore been introduced, which is defined by the ratio between the mean flow velocity
in the median cerebral artery and the mean flow velocity in the internal carotid artery. Criteria for diagnosis
of a stenosis .50% of an intracranial vessel with transcranial Doppler include: (a) segmentary acceleration of
flow velocity; (b) drop in velocity below the stenotic segment; (c) asymmetry; and (d) circumscribed flow
disturbances (turbulence and musical murmur). The transcranial Doppler enables us to assess both components
of self-regulation. The static component is measured by observing changes in flow velocity caused by phar-
macologically induced episodes of hypertension and hypotension. The dynamic component of autoregulation
can be measured using a method devised by Aaslid known as the ‘cuff test’. A very effective and safe device for
measuring cerebral autoregulation is the transient hyperaemic response test. This test is based on the com-
pensatory vasodilatation of the arterioles, which occurs after brief compression of the common carotid.
Csonyka proposed the following formula based on clinical observation for the calculation of cerebral perfusion
pressure: CPP 5 MAP 3 FVd/FVm 1 14. Brain death is defined as the irreversible cessation of all functions of
the whole brain. The clinical criteria are usually considered sufficient to establish a diagnosis of brain death;
however, they might not be sufficient in patients who have been on sedatives or when there are ethical or legal

Correspondence to: Frank A. Rasulo, Spedali Civili University Hospital of Brescia, Institute of Anesthesiology and Intensive Care, Piazzale Spedali Civili, 1 –25125
Brescia, Italy. E-mail: [email protected]; Tel: 139 030 3995 570/764/563
168 F. A. Rasulo et al

controversies. Many authors have demonstrated the existence of a transcranial Doppler pattern, which is
typical of brain death.

Keywords: ULTRASONOGRAPHY DOPPLER TRANSCRANIAL; INTRACRANIAL PRESSURE; REGIONAL

BLOOD FLOW, brain; CRITICAL CARE.

Introduction whose intensity, height and pitch reflect the char-

acteristics of blood flow of the vessel under exam-
Transcranial Doppler (TCD) is an innovative, flex-
ination. In visual mode ‘fast Fourier transform’
ible, accessible tool for the bedside monitoring of
(FFT), also known as spectral analysis, is used to
static and dynamic cerebral flow and treatment
produce a two-dimensional image on the screen.
response. Introduced by Rune Aaslid in 1982 [1], it
The curve is further elaborated to gain several
has become indispensable in clinical practice.
important diagnostic parameters: peak systolic velo-
city, telediastolic velocity, mean velocity, Gosling’s
Physical principles ‘pulsatility index (PI)’ and Pourcelot’s ‘resistance
index (RI)’.
TCD is based on the Doppler principle illustrated
by Christian Andreas Doppler in 1842, which FVs
FVd FVs
FVd
describes the apparent frequency change of a sound PI ¼ ; RI ¼ :
FVm FVs
wave caused by relative movement between the
observer and the sound source [2,3]. When stimu-
lated by an electric current the Doppler probe emits Examination technique
an ultrasound wave, which is reflected by the The main obstacle to ultrasound penetration of the
moving red blood cells it meets: as blood flow skull is bone [6]. Low frequencies, 1–2 MHz, reduce
approaches the probe, frequency of the reflected the attenuation of the ultrasound wave caused by
wave increases while blood flowing away from the bone. TCD also provides the advantage of acoustic
probe produces a lower frequency compared to the windows representing specific points of the skull
original wave. The difference between the frequency where the bone is thin enough to allow ultrasounds
of the original signal and the reflected signal is to penetrate [1,2,7].
known as the Doppler shift and is directly propor- The transtemporal window is sited at the thin
tional to blood flow velocity as expressed below: portion of the temporal bone, between the external
2FtV cos W cantus of the eye and the external acoustic meatus,
F¼ ; immediately above the zygomatic arch. It is the
C most frequently used window and allows investi-
where F is the Doppler shift, Ft is the frequency of gation to be made of the proximal segment (M1) the
the wave emitted, V is the actual velocity, C is the median cerebral artery (MCA), the A1 segment of
velocity of sound in tissue and cos W is the cosine of the anterior cerebral artery (ACA), the posterior
the angle between the insonated vessel and the cerebral artery (PCA) and the final segment of the
direction of the ultrasound wave [2,4,5]. Blood flow internal carotid artery (ICA).
velocity can be estimated by measuring the Doppler By compressing the carotid, it is possible to
shift. The difference between estimated and actual evaluate the patency of the communicating anterior
velocity increases with the angle of incidence W, and posterior arteries. This approach is unsuitable
which means that a better estimate of flow velocity for approximately 10% of patients due to thickened
may be obtained by reducing the angle. bone or osteoporosis. New substances, ultrasound
TCD employs pulsed wave probes featuring a contrast media, have recently been introduced,
single piezoelectric crystal, which is alternately which allow this approach to be used in these
stimulated to produce a signal and silenced to allow patients as well by amplifying the ultrasound beam
the reflected wave to be read. By varying the time reflected.
interval between transmission and reception, it is The transorbital window makes it possible to assess
possible to modulate the depth of insonation and the ophthalmic artery, the carotid siphon and the
examine selectively a specific segment of the cere- contralateral MCA. The probe is placed above
bral vascular tree. TCD provides information on the closed eyelid. This window must not be used
blood flow in both acoustic and visual form. With in patients who have recently undergone surgery
the former an audible acoustic signal is produced to replace the lens as the effects of ultrasound on

r 2008 Copyright European Society of Anaesthesiology, European Journal of Anaesthesiology 25 (Suppl 42): 167–173
 Transcranial Doppler ultrasonography in intensive care 169

artificial lenses are as yet unknown. In addition, the As far as the MCA is concerned, MFVs below
signal power of the Doppler emitted by the trans- 120 cm s21 indicate the presence of a slight reduc-
ducer must be reduced by 20% to avoid damage to tion of the vessel lumen, which cannot show
the retina. up at angiography. Velocities between 120 and
The suboccipital window is sited posteriorly at the 200 cm s21 indicate moderate vasospasm, which
highest portion of the neck. The probe is placed on leads to a reduction of the lumen of between 25%
both sides of the spinal column and must be and 50%, while velocities above 200 cm s21 indi-
orientated towards the median; the ultrasound beam cate serious vasospasm with a reduction of the
penetrates the cranium through the space between lumen exceeding 50%.
the atlas and the base of the skull. This window Nevertheless, an increase in velocity alone is not
makes it possible to evaluate the suboccipital and sufficient to arrive at a diagnosis of vasospasm: a
intracranial portion of both vertebral arteries (VA) condition of hyperaemia also presents with an
and the basilar artery (BA). increase in flow velocity. The Lindegaard Index (LI)
The retromandibular window is not an acoustic [17] has therefore been introduced, which is defined
cranial window as such but represents an extra- by the ratio between the MFV in the MCA and the
cranial approach to assess the distal segment of the MFV in the ICA. Thus, an LI ,3 indicates hyper-
extracranial ICA, immediately before it enters aemia, between 3 and 6 moderate vasospasm and
the carotid foramen. The probe must be placed at .6 serious vasospasm.
the angle of the mandible and orientated cranially. Other parameters, such as velocity increases
The identification of each intracranial vessel is .50% in daily serial examination or the presence of
based on the following elements: (a) velocity and an asymmetry (velocity difference exceeding 50%),
direction; (b) depth of signal capture; (d) possibility can aid in the diagnosis of vasosapsm [18].
of following the vessel its whole length; (e) spatial TCD studies also correlate well with angiography
relationship with other vessels; and (f) response to studies in vasospasm of the BA [19], though less so
homolateral and contralateral carotid compression. for the other arteries at the base of the skull [20].

Clinical applications Stenosis of the intracranial arteries

Two recent publications [8,9] specified the main Criteria for diagnosis of a stenosis .50% of an
fields of clinical application of TCD with an intracranial vessel with TCD include: (a) segmen-
assessment of the advantages and limits of the tary acceleration of flow velocity, (b) drop in velo-
method itself. city below the stenotic segment, (c) asymmetry and
(d) circumscribed flow disturbances (turbulence
Vasospasm and musical murmur) [21–24]. The sensitivity and
diagnostic specificity of the TCD is higher in identi-
Mean flow velocity (MFV) is directly proportional to
fying stenosis of the anterior circulation (carotid
flow and inversely proportional to the section of the
siphon and MCA in its proximal segment M1),
vessel. Any circumstance that leads to a variation of
compared with the posterior circulation (VA, BA at
one of these factors can thus affect mean velocity.
segment P1 of the PCA), which presents more
Physiological or paraphysiological conditions that
anatomic variability and difficulty of insonation.
modify flow velocity are age, Ht, aPCO2 and meta-
Data to establish diagnostic criteria with TCD for
bolic requirement. As for pathological conditions,
stenosis ,50% are insufficient.
the main condition affecting flow velocity is the
vasospasm, the reduction of the lumen of the vessel
following contraction of the smooth muscles of the Intracranial occlusion
wall. Vasospasm is a frequent complication of sub- Doppler diagnosis of occlusion in an intracranial
arachnoid haemorrhage (SAH) [10–15] and has an vessel is possible when no sign of flow can be found
incidence of between 30% and 70%. It has a typical at the normal depth and position, or when its speed
time-span: it is normally absent in the first 48–72 h is significantly reduced and none of the other vessels
after SAH. Onset occurs from day 3 to reach a peak can be heard. There may be increases in speed in
between day 6 and day 12, gradually lessening at other intracranial vessels due to activation of the
15–20 days. The vasospasm often remains clinically compensatory circles. In patients with acute cerebral
silent and the factors that make it symptomatic are ischaemia, the specificity and sensitivity of TCD in
largely unknown [16]. diagnosing MCA occlusions is over 90% [25,26].
Threshold velocities above which vasospasm comes Repeated or continuous TCD monitoring enables us
into place are well defined as regards the MCA, while to follow the evolution of the occlusion and to assess
there is no consensus for the other vessels. possible recanalization in the vessel whether that be

r 2008 Copyright European Society of Anaesthesiology, European Journal of Anaesthesiology 25 (Suppl 42): 167–173
170 F. A. Rasulo et al

spontaneous [27,28] or induced by fibrinolysis. hypotensive stimulus. However, inducing rapid

Comparative studies with angiographs have shown changes in systemic arterial pressure in patients who
excellent diagnostic correlation, even when there are already seriously compromised is not a good
is partial recanalization [29,30]. A recent small idea, so this limits the application of the cuff test.
randomized trial [31] compared 11 patients who A very effective and safe device for measuring
underwent thrombolysis with tissue plasminogen cerebral autoregulation is the transient hyperaemic
activator (t-PA) combined with continuous TCD response test (THRT) [34,35].
monitoring, and 14 patients treated with t-PA, This test is based on the compensatory vasodi-
without TCD monitoring, all of them with acute latation of the arterioles, which occurs after brief
MCA occlusion. The patients who were con- compression of the common carotid. The test
tinuously monitored with TCD had higher levels of involves measuring systolic speed of flow in the
recanalization at 1 h and better outcome at 90 days MCA in base conditions. The common carotid is
compared to those treated with t-PA alone, sug- compressed homo laterally for 10 s, which causes a
gesting that ultrasound may play a role in facil- reduction in CPP. If autoregulation is intact, the
itating the lysis of the thrombus. TCD showed good cerebral arterioles respond to the reduction in CPP
sensitivity and predictability with carotid siphon, through vasodilatation to reduce resistance and
VA and BA [30]. keep the CBF constant. Once the compression is
released, we can see that there is a temporary
Cerebrovascular autoregulation increase in blood flow as CPP acts on a dilatated
Cerebral autoregulation refers to the brain’s intrin- vascular bed.
sic ability to maintain cerebral blood flow (CBF). The increase in flow translates as an increase in
When this system is compromised, patients are at velocity of MCA flow. Autoregulation integrity can
risk of developing ischaemia and cerebral oedema. be checked by calculating the transient hyperaemic
Cerebral autoregulation includes pressure type response ratio (THRR), which is defined as the ratio
regulation, which means that changes in CBF are between the velocity of systolic flow during the
kept to a minimum even when cerebral perfusion hyperaemic phase (two cycles after the compression
pressure (CPP) varies, as long as the variations are release excluding the very first cycle) and the velo-
within 50 and 150 mmHg, and vasomotor react- city of basic systolic flow (five cycles before com-
ivity, which is the response of the cerebral circle to pression). The normal THRR range is between
changes in aPCO2 and aPO2. 1.105 and 1.29 (average (95% CI) 1.2 (1.17–1.24)).
Pressure regulation can be subdivided into THRR is said to be a qualitative autoregulation
dynamic regulation, involving a rapid response on indicator. The strength and duration of the carotid
the part of CBF to changes in arterial pressure or its compression are the two variables that make THRR
pulsatile nature, and static regulation, involving a unreliable as a quantitative indicator. In the litera-
rapid response on the part of CBF to slow changes ture there is disagreement about how long the
in average arterial pressure [32] (mean arterial compression phase needs to be to get the maximum
pressure (MAP)). hyperaemic response. Some authors consider 5 s [34]
The TCD enables us to assess both components of whereas others consider 10 s [36]. As far as the
self-regulation. The static component is measured strength of the compression is concerned, it needs to
by observing changes in flow velocity caused by be sufficient to cause a reduction in cerebral flow of
pharmacologically induced episodes of hypertension at least 40% [36].
and hypotension. In this way, the static rate of
autoregulation can be measured using the ratio Estimating intracranial pressure
between percentage variation of resistance and per- Increased ICP causes variation in CBF speed, which
centage variation of MAP. A value of 1 indicates translates as changes in the shape of the TCD wave
that autoregulation is intact, whereas 0 tells us that produced [37,38]. Differences in the Doppler wave
autoregulation has stopped [33]. can be quantified by calculating the pulsatility rate
The dynamic component of autoregulation can be (PR) and resistance rate.
measured using a method devised by Aaslid known Various authors have demonstrated how PR
as the ‘cuff test’. Arterial hypotension is induced by increases exponentially when there is intracranial
placing thigh cuffs on each thigh and tightening hypertension. In cranial trauma patients, raised PR
them to 50 mmHg above the PAS for 3 min and is an indicator predicting worse outcome [39]. In
then quickly releasing them. The dynamic rate of the second half of the 1980s, Klingelhofer [40,41]
regulation, dRoR, whose normal value is 20% s21 is showed there was good correlation between ICP and
thus calculated, indicating how quickly the velocity the MAP 3 RI/FVm ratio (where MAP is average
of cerebral flow returns to its starting level after the arterial pressure and FVm the average velocity of

r 2008 Copyright European Society of Anaesthesiology, European Journal of Anaesthesiology 25 (Suppl 42): 167–173
 Transcranial Doppler ultrasonography in intensive care 171

flow). However, many different factors can affect If we have MAP and FV for a single cardiac cycle in
levels of pulsatility and resistance including hemo- a single cartesian axis, we can calculate a rate of
dynamic, respiratory and haematological factors as linear regression. Once we have the linear regression
well as vascular and tissue compliance. This is why rate, we can extrapolate the AP value corresponding
indicators such as these cannot be used for the early to zero flow. This value corresponds to EDP.
identification of patients at risk for the development aPCO2 variations have the opposite effect on ICP
of endocranial hypertension. and CCP. Increased CO2 when autoregulation is
In 1986, Aaslid [42] applied the Fourier analysis preserved leads to vasodilatation, which, on the one
to the Doppler and arterial pressure waves and came hand, causes an increase in cerebral blood volume and
up with the following formula for calculating CPP: therefore ICP, and, on the other, leads to a reduction
CPP5AP1 3 FVm/FV1, where AP1 is the ampli- in vascular tone and therefore reduction in CCP.
tude of the first peak of the AP wave and FV1 is the Weyland and colleagues [46] questioned what would
amplitude of the first peak of the Doppler wave. happen if two Starling resistors were placed in
More recently, Czsonyka and colleagues [43] pro- sequence, one at the arteriole level so mainly influ-
posed the following formula based on clinical enced by CCP, and the other at vein level so influ-
observation: CPP5MAP 3 FVd/FVm114, where enced mainly by ICP. According to the author, if
FVd is the velocity of diastolic flow. In a group of there is no endocranial hypertension, the eCPP is
25 patients with serious cranial trauma, absolute mainly determined by the arteriole resistor. Tradi-
error was less than 10 mmHg in 81% of the cases tionally, we have always thought of CPP as the dif-
and less than 5 in 50% of the cases. Using this ference between MAP and ICP. However, it is also
method, a prototype has been put forward, which likely that real CPP is not determined by intracranial
allows for continuous and bilateral CPP measure- pressure but by CCP. In a study of 70 patients
ment (Neuro Q TM Deltex Ltd, Chichester, UK). with serious trauma, these showed that in 51% of
This technique, based on critical closing pressure cases CPP-ICP underestimates eCPP by at least
(CCP), has proved to be fairly reliable in predicting 19.8 mmHg [47]. However, there is no evidence to
CPP values, but it is not sufficiently precise where suggest that the concept of eCPP is superior to that
ICP is concerned in that it does not differentiate of CPP in terms of outcome. The mechanism by
between the effects on CPP, which are caused by which the difference between EDP and ICP can be a
the increase in ICP and those produced by an negative also needs to be clarified. It may be due to
increase in cerebrovascular resistence, especially in vasoparalysis or marked dilatation at rest because of
patients with intact autoregulation or in those who hypercapnea, hypertension or hypoxia. The literature
have intra- or extracranial stenosis. reveals contrasting opinions on this.

Transcranial Doppler and ‘effective

downstream pressure’ Transcranial Doppler and brain death
There has been a recent return to the concept of Brain death is defined as the irreversible cessation of
‘effective downstream pressure (EDP)’ and CCP. The all functions of the whole brain [48]. The clinical
concept of CCP was first mentioned in the 1950s by criteria are usually considered sufficient to establish a
Burton [44], who defined it as the minimum diagnosis of brain death; however, they might not be
transmural pressure (MAP-ICP) under which blood sufficient in patients who have been on sedatives or
flow stops and the vessel collapses. Critical closing when there are ethical or legal controversies. In these
pressure corresponds to vasomotor tone [45]. circumstances, it might be useful to have tests, which
When there is equilibrium (zero flow), trans- could confirm a diagnosis of brain death. The use of
mural pressure is equal to the ratio between the wall tests is also recommended in patients with severe
tension, which is given by the vasomotor tone, and facial trauma, in patients with pre-existing altera-
the radius of the vessel. A reduction in arterial tions in pupillary diameter and in PBCO patients
pressure, an increase in ICP or in vasomotor tone who normally have high aPCO2 levels.
can modify the balance between transmural pressure The most frequently used test is EEG, even
and wall tension, leading to the collapse of the though it gives very little information about how
vessel. The sum of CPP plus ICP gives us EDP. the brain stem is, and it is not always technically
When EDP equals MAP, flow is zero. The difference possible in ICUs to conduct the test properly.
between MAP, and EDP is effective CPP (eCPP), Angiography is more suitable for confirming brain
which CBF depends on. death. However, it remains an invasive test, which
EDP values can be found by analysing the beat- requires the use of contrast and therefore the patient
to-beat relationship between pressure and CBF. needs to be taken out of Intensive Care. The same

r 2008 Copyright European Society of Anaesthesiology, European Journal of Anaesthesiology 25 (Suppl 42): 167–173
172 F. A. Rasulo et al

goes for computed tomography and any techniques References

that use radioisotopes.
TCD is a test that is relatively simple to perform, 1. Aaslid R, Markwalder TM, Nornes H. Noninvasive
transcranial Doppler ultrasound recording of flow
cheap and that can be done at the patient’s bedside.
velocity in basal cerebral arteries. J Neurosurg 1982; 57:
Many authors have demonstrated the existence of a 769–774.
TCD pattern, which is typical of brain death [49]. 2. McCartney JP, Thomas-Lukes KM, Gomez CR. Handbook
The increased ICP, which leads to the arrest of of Transcranial Doppler. New York, USA: Springer, 1997.
the brain circle, first causes a reduction in diastolic 3. Aaslid R. Transcranial Doppler Sonography. New York, USA:
flow velocity, which becomes equal to 0 when ICP Springer, 1986.
approximates diastolic arterial pressure. If ICP 4. Manno EM. Transcranial Doppler ultrasonography
continues to increase, diastolic flow reappears but in in the neurocritical care unit. Crit Care Clin 1997; 13:
the opposite direction (reverberating or oscillating 79–104.
flow), indicating a retrograde flow during the dia- 5. Eicke BM, Tegeler CH, Dalley G et al. Angle correction in
stolic phase of the cardiac cycle. When endocranial transcranial Doppler sonography. J Neuroimaging 1994; 4:
hypertension is sufficiently high as to cause the 29–33.
6. Grolimund P. Transmission of Ultrasound through
arrest of CBF, we observe brief systolic spikes fol- Temporal Bone. In: Aaslid R, ed. Transcranial Doppler
lowed by the complete disappearance of the Doppler Sonography. New York, USA: Springer, 1986.
signal. A reverberating flow and systolic spikes are 7. Babikian Vl, Wechsker Lr. Transcranial Doppler Ultrasono-
considered conclusive of cerebral circle arrest, graphy. Butterworth-Heinemann 1999.
because it means that both the anterior and pos- 8. Babikian VL, Feldmann E, Wechsler LR et al. Transcranial
terior as well as the lateral ones are affected. We Doppler ultrasonography: year 2000 update. J Neuroima-
cannot exclude the possibility that the acoustic ging 2000; 10: 101–115.
window is not sufficient, so if there is no signal 9. Sloan MA, Alexandrov AV, Tegeler CH et al. Assessment:
whatsoever in any of the vessels at the base of the transcranial Doppler ultrasonography: report of the
cranium, this is considered indicative of brain death Therapeutics and Technology Assessment Subcommittee
only when previous examination showed that flow of the American Academy of Neurology. Neurology 2004;
62: 1468–1481.
was present. To exclude temporary arrest because of 10. Aaslid R, Huber P, Nornes H. Evaluation of cerebrovas-
hypotension, systolic arterial pressure must not be cular spasm with transcranial Doppler ultrasound.
less than 70 mmHg and flow patterns need to have J Neurosurg 1984; 60: 37–41.
been monitored during at least two previous 11. Vora YY, Suarez-Almazor M, Steinke DE et al. Role
examinations at least 30 min apart. of transcranial Doppler monitoring in the diagnosis
The sensitivity of TCD is 96.5% and its speci- of cerebral vasospasm after subarachnoid hemorrhage.
ficity is 100% in the diagnosis of brain death. Neurosurgery 1999; 44: 1237–1247.
However, it should be used to complement rather 12. Sloan MA, Haley Jr EC, Kassell NF et al. Sensitivity and
than substitute careful clinical evaluation. It is also specificity of transcranial Doppler ultrasonography in the
important to stress that central nervous system diagnosis of vasospasm following subarachnoid hemor-
depressor drugs do not influence the diagnosis of rhage. Neurology 1989; 39: 1514–1518.
13. Lysakowski C, Walder B, Costanza MC et al. Transcranial
brain death with TCD.
Doppler versus angiography in patients with vasospasm
due to a ruptured cerebral aneurysm: a systematic review.
Patent oval foramen Stroke 2001; 32: 2292–2298.
14. Laumer R, Steinmeier R, Gonner F et al. Cerebral
Paradoxical embolism through the patent oval hemodynamics in subarachnoid hemorrhage evaluated by
foramen is a possible cause of stroke in young transcranial Doppler sonography. Part 1. Reliability of
patients. TCD enables us to diagnose the presence of flow velocities in clinical management. Neurosurgery 1993;
a right–left cardiac shunt with almost 100% con- 33: 1–8.
cordance if we compare it with a transoesophageal 15. Steinmeier R, Laumer R, Bondar I et al. Cerebral
ecograph. In all, 9 cm3 of physiological solution hemodynamics in subarachnoid hemorrhage evaluated by
is injected into one of the large forearm veins transcranial Doppler sonography. Part 2. Pulsatility
through a three-way tap, 1 cm3 of air. If there is indices: normal reference values and characteristics in
subarachnoid hemorrhage. Neurosurgery 1993; 33: 10–18.
right–left shunt, we can see the gaseous emboli 16. Grosset DG, Straiton J, McDonald I et al. Use of
moving in the MCA 5–15 s after the injection. If transcranial Doppler sonography to predict development
the result of the basic test is negative, we can of a delayed ischemic deficit after subarachnoid hemor-
increase the sensitivity of the test by performing it rhage. J Neurosurg 1993; 78: 183–187.
while conducting a Valsalva’s [50] manoeuvre. Extra- 17. Lindegaard KF. The role of transcranial Doppler in the
cardiac shunt (e.g. intrapulmonary) is responsible for management of patients with subarachnoid haemorrhage –
false positives. a review. Acta Neurochir 1999; 72: 59–71.

r 2008 Copyright European Society of Anaesthesiology, European Journal of Anaesthesiology 25 (Suppl 42): 167–173
 Transcranial Doppler ultrasonography in intensive care 173

18. Grosset DG, Straiton J, du Trevou M et al. Prediction of 34. Giller GA. A bedside test for cerebral autoregulation using
symptomatic vasospasm after subarachnoid hemorrhage by transcranial Doppler ultrasound. Acta Neurochirurgica 1991;
rapidly increasing transcranial Doppler velocity and 108: 7–14.
cerebral blood flow changes. Stroke 1992; 23: 674–679. 35. Smielewsky P, Czosnyka M, Kirkpatrik P et al. Evaluation
19. Sloan MA, Burch CM, Wozniak MA et al. Transcranial of the transient hyperemic response test in head injured
Doppler detection of vertebrobasilar vasospasm following patients. J Nerurosug 1997; 86: 773–778.
subarachnoid hemorrhage. Stroke 1994; 25: 2187–2197. 36. Cavill G, Simpson Ej, Mahajan RP. Factor affecting
20. Lennihan L, Petty GW, Fink E et al. Transcranial Doppler assessment of cerebral autoregulation using the transient
detection of anterior cerebral artery vasospasm. J Neurol hyperaemic response test. Br J Anesth 1998; 81: 317–321.
Neurosurg Psychiatry 1993; 56: 906–909. 37. McQuire JC, Sutcliffe JC, Coats TJ. Early changes in
21. Rorick MB, Nichols FT, Adams RJ. Transcranial Doppler middle cerebral artery blood flow velocity after head
correlation with angiography in detection of intracranial injury. J Neurosurg 1998; 89: 526–532.
stenosis. Stroke 1994; 25: 1931–1934. 38. Ursino M, Giulioni M, Lodi CA. Relationships among
22. de Bray JM, Joseph PA, Jeanvoine H et al. Transcranial cerebral perfusion pressure, autoregulation, and transcra-
Doppler evaluation of middle cerebral artery stenosis. nial Doppler waveform: a modeling study. J Neurosurg
J Ultrasound Med 1988; 7: 611–616. 1998; 89: 255–266.
23. Schwarze JJ, Babikian V, DeWitt LD et al. Longitudinal 39. Chan KH, Dearden NM, Miller JD et al. Transcranial
monitoring of intracranial arterial stenoses with transcra- Doppler waveform differences in hyperemic and non-
nial Doppler ultrasonography. J Neuroimaging 1994; 4: hyperemic patients after severe head injury. Surg Neurol
182–187. 1992; 38: 433–436.
24. Wilterdink JL, Feldmann E, Furie KL et al. Transcranial 40. Klingelhofer J, Conrad B, Benecke R et al. Evaluation of
Doppler ultrasound battery reliably identifies severe intracranial pressure from transcranial Doppler studies in
internal carotid artery stenosis. Stroke 1997; 28: 133–136. cerebral disease. J Neurol 1988; 235: 159–162.
25. Zanette EM, Fieschi C, Bozzao L et al. Comparison of 41. Klingelhofer J, Conrad B, Benecke R et al. Intracranial
cerebral angiography and transcranial Doppler sonography flow patterns at increasing intracranial pressure. Klin
in acute stroke. Stroke 1989; 20: 899–903. Wochenschr 1987; 65: 542–545.
26. Camerlingo M, Casto L, Censori B et al. Transcranial 42 Aaslid R. Intracranial pressure IV. New York: Springer,
Doppler in acute ischemic stroke of the middle cerebral 1986: pp. 226–229.
artery territotories. Acta Neurol Scand 1993; 88: 108–111. 43. Czosnyka M, Matta BF, Smielewski P et al. Cerebral
27. Kushner MJ, Zanette EM, Bastianello S et al. Transcranial perfusion pressure in headinjured patients: a noninvasive
Doppler in acute hemispheric brain infarction. Neurology assessment using transcranial Doppler ultrasonography.
1991; 41: 109–113. J Neurosurg 1998; 88: 802–808.
28. Toni D, Fiorelli M, Zanette EM et al. Early spontaneous 44. Burton Ac. On the physical equilibrium of small blood
improvement and deterioration of ischemic stroke vessels. Am J Physiol 1951; 164: 319–329.
patients. A serial study with transcranial Doppler 45. Richards HK, Czosnyka M, Pickard JD. Assessment of
ultrasonography. Stroke 1998; 29: 1144–1148. critical closing pressure in the cerebral circulation as a
29. Demchuk AM, Christou I, Wein TH et al. Accuracy and measure of cerebrovascular tone. Acta Neurochir 1999; 141:
criteria for localizing arterial occlusion with transcranial 1221–1227, discussion 1226–1227.
Doppler. J Neuroimaging 2000; 10: 1–12. 46. Weyland A, Buhre W, Grund S et al. Cerebrovascular tone
30. Burgin WS, Malkoff M, Felberg RA et al. Transcranial rather than intracranial pressure determines the effective
Doppler ultrasound criteria for recanalization after downstream pressure of the cerebral circulation in the
thrombolysis for middle cerebral artery stroke. Stroke absence of intracranial hypertension. J NeurosurgAnesthesiol
2000; 31: 1128–1132. 2000; 12: 210–216.
31. Eggers J, Koch B, Meyer K et al. Effect of ultrasound on 47. Thees C, Scholz M, Schaller MDC et al. Relationship
thrombolysis of middle cerebral artery occlusion. Ann between intracranial pressure and critical closing pressure
Neurol 2003; 53: 797–800. in patients with neurotrauma. Anesthesiology 2002; 96:
32. Lang Ew, Diehl R, Mehdorn M. Cerebral autoregula- 595–599.
tion testing after subarachnoid hemorrhage: the phase 48. Wijdicks EF. The diagnosis of brain death. N Engl J Med
relationship between arterial blood pressure and 2001; 344: 1215–1221.
cerebral blood flow velocity. Critical Care Med 2001; 29: 49. Ropper AH, Kehne SM, Wechsler L. Transcranial Doppler
158–163. in brain death. Neurology 1987; 37: 1733–1735.
33. Czosnyka M, Kirkpatrik P, Pickard JD. Multimodal 50. Droste DW, Lakemeier S, Wichter T et al. Optimizing the
monitoring and assessment of cerebral haemodynamic technique of contrast transcranial Doppler ultrasound in
reserve after severe head injury. Cerebrovasc Brain Metab Rev the detection of right-to-left shunts. Stroke 2002; 33:
1996; 8: 273–295. 221–226.

r 2008 Copyright European Society of Anaesthesiology, European Journal of Anaesthesiology 25 (Suppl 42): 167–173