Neuroimaging in REVIEW ARTICLE

Acute Stroke

C O N T I N U UM A U D I O
I NT E R V I E W A V A I L AB L E
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

ONLINE
By Bijoy K. Menon, MD, DM, MSc, FRCPC
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

ABSTRACT
PURPOSE OF REVIEW: This article describes how imaging can be used by
physicians in diagnosing, determining prognosis, and making appropriate
treatment decisions in a timely manner in patients with acute stroke.

RECENT FINDINGS:Advances in acute stroke treatment, including the use of

tbw+JMEi5ZgHD on 07/29/2023

endovascular thrombectomy in patients with large vessel occlusion and,

more recently, of IV thrombolysis in an extended time window, have
resulted in a paradigm shift in how imaging is used in patients with acute
stroke. This paradigm shift, combined with the understanding that “time is
brain,” means that imaging must be fast, reliable, and available around the
clock for physicians to make appropriate clinical decisions. CT has
therefore become the primary imaging modality of choice. Recognition of a
large vessel occlusion using CT angiography has become essential in
identifying patients for endovascular thrombectomy, and techniques such
as imaging collaterals on CT angiography or measuring blood flow to
predict tissue fate using CT perfusion have become useful tools in
selecting patients for acute stroke therapy. Understanding the use of these
imaging modalities and techniques in dealing with an emergency such as
acute stroke has therefore become more important than ever for CITE AS:
CONTINUUM (MINNEAP MINN)
physicians treating patients with acute stroke. 2020;26(2, CEREBROVASCULAR
DISEASE):287–309.
SUMMARY: Imaging the brain and the blood vessels supplying it using modern
Address correspondence to
tools and techniques is a key step in understanding the pathophysiology of
Dr Bijoy Menon, 1079 29th St NW,
acute stroke and making appropriate and timely clinical decisions. Calgary, AB, Canada T3H4J2,
[email protected].

RELATIONSHIP DISCLOSURE:
Dr Menon has received
INTRODUCTION research/grant support from

A
cute stroke is an emergency. Treatment decisions in acute stroke Alberta Innovates, the Canadian
are, however, made after carefully considering clinical information Institutes of Health Research
(363959), and the Heart and
and assessing available imaging. As time is of the essence in the Stroke Foundation of Canada.
management of patients with acute stroke, this clinical assessment Dr Menon has held stock in
and imaging interpretation must happen quickly.1,2 Balancing the Circle Neurovascular
Imaging Inc.
need for rigor in evaluating these patients while at the same time recognizing the
need for speed in such evaluations is a trade-off many physicians find difficult to UNLABELED USE OF
PRODUCTS/INVESTIGATIONAL
make. However, to achieve optimal outcomes for their patients, physicians must USE DISCLOSURE:
be able to seek and acutely receive the clinical and imaging information that is Dr Menon reports no disclosure.
absolutely essential in making treatment decisions.
This article discusses how to use the available clinical and appropriate imaging © 2020 American Academy
information to make timely treatment decisions in patients with acute stroke. A of Neurology.

CONTINUUMJOURNAL.COM 287

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

case vignette is used to help readers better understand this process. The first step
in assessing a patient with suspected acute stroke is to check the patient’s vital
signs (airway stability, heart rate, and blood pressure). If the patient’s vital signs
are stable, imaging should be arranged immediately. In parallel, without delaying
the patient’s transport to the CT scanner for imaging, the physician should get a
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

brief history from a witness or the paramedics while doing a focused neurologic
examination (CASE 3-1A). History taking should focus on obtaining the following
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

information as quickly as possible:

u Time when the patient was last known well

u Nature and severity of symptoms
u History of previous strokes or cardiovascular disease
u History of vascular risk factors (diabetes mellitus, hypertension, dyslipidemia, or
atrial fibrillation)
u Medication history, in particular, ascertaining if the patient is taking any antiplatelet
tbw+JMEi5ZgHD on 07/29/2023

agents or anticoagulants
u History of recent surgery or bleeding
u Any serious comorbidity (eg, cancer, dementia, chronic renal disease)

The focused neurologic examination is done in parallel with the history taking,
primarily to assess the severity of clinical deficits (eg, hemiplegia versus
hemiparesis) and to look for the presence or absence of cortical signs (aphasia,
hemianopia, or hemineglect). The National Institutes of Health Stroke Scale
(NIHSS; nihstrokescale.org) is often used as a structured assessment tool in
patients with acute stroke.3

CASE 3-1A A 72-year-old man was at home having dinner with his wife when she
noticed that he was slurring his words. Over the next minute, she
noticed that he was unable to move his right side. She immediately
called 911. Paramedics transported the patient to the nearest primary
stroke center. When he arrived at the hospital, he was 3 hours from stroke
symptom onset.
The patient’s symptoms were severe. On neurologic assessment, his
National Institutes of Health Stroke Scale score was 15. He had
expressive aphasia, right hemiplegia with ipsilateral upper motor
neuron facial paresis, and gaze preference toward the left. He had a
history of diabetes mellitus and hypertension and was taking aspirin
81 mg once a day. He had no history of any recent surgery or bleeding
and no serious comorbidities.

COMMENT Time is of the essence when diagnosing stroke; therefore, patient

assessment in acute stroke should be focused but fast and without a
delay in imaging.

288 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 IMAGING
Imaging is used in acute stroke to help answer the following questions with
regard to diagnosis, prognosis, and treatment selection:

u Diagnosis: Does the patient have a stroke or not? Is the stroke ischemic or hemorrhagic?
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

u Prognosis: What is the size of the stroke? Is the stroke already so big that the patient
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

is likely to have a poor prognosis?

u Treatment selection: Would this patient with ischemic stroke benefit from thrombolysis,
endovascular treatment, or a combination of both? What is the risk of bleeding
with thrombolysis?

Answers to these questions help physicians make treatment decisions. CT and

MRI are the two imaging modalities used in acute stroke. Although MRI is
considered a more sophisticated imaging tool than CT, CT is the workhorse of
acute stroke imaging.4,5 The primary reasons for the almost universal use of CT in
acute stroke are the speed and ease of acquisition, 24/7 availability, lower cost,
tbw+JMEi5ZgHD on 07/29/2023

and relative absence of contraindications when compared to MRI. CT modalities

include noncontrast CT, CT angiography (CTA), and CT perfusion (CTP). These
three imaging modalities provide different diagnostic, prognostic, and treatment
selection information.

Noncontrast CT
Noncontrast CT is the first imaging obtained in most patients with stroke for
rapid assessment (TABLE 3-1). It provides images of the skull, brain parenchyma,
ventricles, and adjacent structures. Noncontrast CT in patients with acute stroke
is used to rule out a hemorrhagic stroke and to identify imaging features that may
suggest the presence of an ischemic stroke (FIGURE 3-1). Ischemic brain lesions
are dark (hypoattenuated) on noncontrast CT when compared to hemorrhages,
which are bright (hyperattenuated).

HEMORRHAGE. The most common form of hemorrhage in patients with acute

stroke is hemorrhage within the brain parenchyma (intracerebral hemorrhage).
Hemorrhages may be classified based on their anatomic location: intracerebral
(superficial versus deep), intraventricular, subarachnoid, subdural, or

Imaging Features on Noncontrast CT to Rapidly Assess Patients Presenting TABLE 3-1

With Acute Stroke

◆ Any intracerebral, intraventricular, subarachnoid, subdural, or extradural hemorrhage

◆ Eye deviation
◆ Hyperdense artery sign
◆ Early ischemic changes (obscuration of lentiform nucleus, insular ribbon sign, loss of
gray-white differentiation of surface cortex)
◆ Alberta Stroke Program Early CT Score (ASPECTS)a (good 8-10, intermediate 5-7, poor 0-4)

a
ASPECTS assessment does not apply to posterior circulation strokes.

CONTINUUMJOURNAL.COM 289

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

KEY POINTS

● As time is of the essence

in the management of
patients with acute stroke,
clinical assessment and
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

imaging interpretation must

happen quickly.
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

● Imaging is used in acute

stroke to help determine
diagnosis, prognosis, and
appropriate treatment
selection.

● CT is the workhorse of
acute stroke imaging
because of its speed and
ease of acquisition, 24/7
tbw+JMEi5ZgHD on 07/29/2023

availability, lower cost, and FIGURE 3-1

relative absence of Noncontrast CT in patients presenting with acute stroke. A, Axial noncontrast CT showing
contraindications when infarction in the left cerebral hemisphere. B, Axial noncontrast CT showing intracerebral
compared to MRI. hemorrhage in the left temporal region with extension into the adjacent subarachnoid space.

● The primary purpose of

noncontrast CT in patients
with acute stroke is to rule extradural (FIGURE 3-2).6 This anatomic classification helps determine the likely
out a hemorrhagic stroke etiology of the hemorrhage (TABLE 3-2).
and to identify imaging Noncontrast CT is also useful in determining the volume of the hemorrhage.
features that may suggest
the presence of an ischemic The method used most often was popularized by Kothari and colleagues7 and
stroke. modified from the formula for determining the volume of an ellipsoid (ie, 4/3 π x
[A/2] x [B/2] x [C/2], where A, B, and C are the three diameters of the ellipsoid).
● The classic radiologic The formula reduces to (A x B x C)/2 if π is approximated to the number 3.
signs of early ischemic
change seen on noncontrast
Therefore, on noncontrast CT,
CT are obscuration of the A is the longest measurement of
lentiform nucleus, the hemorrhage in the axial plane, B
insular ribbon sign (loss of is the longest measurement of
gray-white matter
hemorrhage diameter
differentiation at the insula),
and the cortical ribbon sign perpendicular to A in the same
(loss of gray-white matter or different axial plane, and C is
differentiation at the surface longest measurement of
cortex).
hemorrhage in the vertical
plane (determined simply by
multiplying the number of CT
slices with hemorrhage by the
slice thickness). It is
recommended that these
measurements be made in
centimeters to obtain a volume
in milliliters. A baseline
hemorrhage volume of 60 mL or
higher is often considered as being FIGURE 3-2
Locations and types of intracranial hemorrhage
associated with poor prognosis;
that may be seen on noncontrast CT.
although such prognostication Modified with permission from Al-Shahi Salman R, et al,
may need to take into account BMJ.6 © 2009 British Medical Journal Publishing Group.

290 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 other variables such as the patient’s age and other existing comorbidities.8,9 More
recently identified markers of poor prognosis with intracerebral hemorrhage
seen on noncontrast CT include the blend sign (a hemorrhage with hyperdense
and hypodense regions), the black hole sign (a hypodense region surrounded by
a hyperdense region not connected with the adjacent brain tissue), and the island
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

sign (at least three scattered small hemorrhages separate from the main
hemorrhage) (FIGURE 3-3).10
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

ISCHEMIA. Severe cerebral ischemia (blood flow <10 mL/100 g/min) causes brain
tissue water to shift from the extracellular to the intracellular space, thus resulting
in hypoattenuation on noncontrast CT.11–13 Experiments in animal models show
that a 1% increase in brain tissue water content is associated with a 1.8–Hounsfield
unit decrease in attenuation.12 A Hounsfield unit is a measure of attenuation on
noncontrast CT. In noncontrast CT of a normal brain, gray matter appears brighter
than white matter. With severe ischemia, gray matter becomes edematous, loses
signal, and starts to become darker on noncontrast CT. This results in blurring and
tbw+JMEi5ZgHD on 07/29/2023

then loss of normal gray-white matter differentiation. These changes, called early
ischemic changes, can be seen as early as 1 hour after stroke onset.14 These changes
are often considered as evidence of irreversibly infarcted brain.

RADIOLOGIC SIGNS OF EARLY ISCHEMIC CHANGES. The classic radiologic signs of

early ischemic changes seen on noncontrast CT are obscuration of the lentiform
nucleus, the insular ribbon sign (loss of gray-white matter differentiation at the
insula), and the cortical ribbon sign (loss of gray-white matter differentiation at
the surface cortex) (FIGURE 3-4).

EXTENT OF ISCHEMIA. To select patients for acute therapies such as thrombolysis

and endovascular treatment, physicians must determine if brain tissue exists
that is still alive and can therefore be saved. This concept of assessing brain tissue
to determine the extent of brain that is likely irreversibly infarcted versus the
extent of brain that is still salvageable is known as mismatch. Imaging is used to
assess this mismatch; on noncontrast CT, mismatch is assessed by determining

Anatomic Classification of Intracranial Hemorrhage and Corresponding TABLE 3-2

Etiology

Location Etiology

Intraparenchymal

Lobar Amyloid angiopathy, bleeding diathesis, hypertension, vascular

malformations, tumors, venous thrombosis, vasculitis

Deep Hypertension, tumors, moyamoya syndrome

Intraventricular Hypertension, tumors

Subarachnoid Aneurysms, amyloid angiopathy, trauma

Subdural Trauma, bleeding diathesis

Extradural Trauma

CONTINUUMJOURNAL.COM 291

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

FIGURE 3-3
tbw+JMEi5ZgHD on 07/29/2023

Imaging characteristics of intracerebral hemorrhage associated with poor prognosis. A, Axial

noncontrast CT showing the island sign (ie, at least three scattered small hemorrhages
separate from the main hemorrhage) (white arrows), whereas the hemorrhage has
hyperdense and hypodense regions (the blend sign) (black arrow). B, Axial noncontrast CT
showing the black hole sign (a hypodense region surrounded by a hyperdense region not
connected with the adjacent brain tissue) (arrow).
Reprinted with permission from Sporns PB, et al, Stroke.10 © 2018 Wolters Kluwer Health.

the extent of early ischemic changes in ischemic brain. Thus, if a patient presents
with symptoms and signs that suggest involvement of the left cerebral
hemisphere, assessing the extent of early ischemic changes in the left cerebral
hemisphere helps determine if brain tissue exists that can still be saved with
treatment.15 The two most widely used methods to assess the extent of early
ischemic changes in brain supplied by the middle cerebral artery (MCA) are the
one-third MCA rule and the Alberta Stroke
Program Early CT Score (ASPECTS).
The one-third MCA rule assesses the
extent of early ischemic changes in the
area of the brain supplied by the MCA. If
more than one-third of the MCA vascular
territory has early ischemic changes, the
benefit of thrombolysis is limited and risk
of postthrombolysis intracerebral
hemorrhage is higher.16 Although the
one-third MCA rule is a useful marker of
prognosis, statistical evidence of its
usefulness in selecting patients for
treatment is limited. Moreover, it lacks
reliability in real-world situations.
The ASPECTS method was developed
to address the limitations of the one-third FIGURE 3-4
MCA rule and to assess early ischemic Axial noncontrast CT showing lentiform
nucleus obscuration (arrow) and the
changes systematically. The ASPECTS is a insular ribbon sign (star). Some cortical
10-point ordinal scale that measures regions still retain gray-white matter
parenchymal hypoattenuation or loss of differentiation.

292 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 gray-white matter differentiation KEY POINT
in the brain territory supplied by
● The two most widely used
the MCA (FIGURE 3-5).17 It methods to assess the
assigns equal weight to smaller extent of early ischemic
subcortical structures (such as changes in brain supplied by
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

the caudate, lentiform, and the middle cerebral artery

are the one-third middle
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

internal capsule) compared with

cerebral artery rule and the
the insula and the six cortical Alberta Stroke Program
regions. Early ischemic changes Early CT Score (ASPECTS).
are assessed on the axial plane of
noncontrast CT at the level of the
caudate head (caudate head,
lentiform nucleus, internal
capsule, and M1 to M3 regions;
called the ganglionic level) and at
a level above the caudate head
tbw+JMEi5ZgHD on 07/29/2023

FIGURE 3-5
(M4 to M6 regions; called the The Alberta Stroke Program Early CT Score
supraganglionic level). The (ASPECTS) template applied to noncontrast CT.
caudate nucleus involvement is M1 through M6 are six distinct cortical regions in
the middle cerebral artery vascular territory.
assessed at both the ganglionic C = caudate nucleus; I = insula; IC = internal capsule;
level (caudate head) and the L = lentiform nucleus.
supraganglionic level (caudate Reprinted from Puetz V, et al, Int J Stroke.14 © 2009 SAGE
Publications.
body and tail).
ASPECTS is best assessed on
axial noncontrast CT with at least 5-mm slice thickness. A region should only be
considered as involved if early ischemic changes are detected on at least two
adjacent slices. If an ASPECTS region has early ischemic changes, that region is
assigned a score of 0; the region is given a score of 1 if no early ischemic changes
are detected. The scores of the regions are then added to get the total ASPECTS. In
a patient with symptoms and signs of ischemic cerebral hemispheric involvement,
a low score (eg, 0) suggests extensive early ischemic changes, poor prognosis, and
no brain to save, whereas a score of 10 suggests the patient has no early ischemic
changes and a good prognosis, potentially with successful recanalization.18
ASPECTS assessment is affected by CT quality.19 It is therefore important that
the window levels of the noncontrast CT are adjusted so that gray-white matter
differentiation is visualized well in the cerebral hemisphere considered normal
(FIGURE 3-6A). Patient motion also affects the assessment of early ischemic
changes at the regional level (FIGURE 3-6B). Regional involvement should also be
scored cautiously if bone or metal artifacts are present. If in doubt, the region
should not be considered abnormal.

SUBACUTE ISCHEMIC CHANGES ON NONCONTRAST CT. The degree of

hypoattenuation on noncontrast CT increases with time and with increasing
severity of ischemia.13 A patient who presents many hours after stroke symptom
onset may have cortical or deep gray matter regions that look darker than
normal-appearing white matter. These changes are called “subacute” changes
rather than early ischemic changes. These subacute-appearing regions on
noncontrast CT can be seen not only in patients who present late but also in
patients who present earlier but have more severe ischemia. To assess for
subacute changes, it is best to compare the hypoattenuation in the concerning

CONTINUUMJOURNAL.COM 293

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0
tbw+JMEi5ZgHD on 07/29/2023

FIGURE 3-6
Variables affecting the Alberta Stroke Program Early CT Score (ASPECTS) assessment on
noncontrast CT for a patient with ischemic stroke of the left hemisphere and an ASPECTS
of 9. The only area affected is a small part of the left lentiform nucleus. A, Window level
settings and how they affect noncontrast CT interpretation. All images are the same slice
from the same patient. The goal of window setting in acute ischemic stroke assessment is to
get the best discrimination between gray and white matter. B, The effect of patient motion on
interpretation of ASPECTS. The scan without motion artifact shows that the left caudate
nucleus and the left M1 regions are spared unlike in the scan with motion.
C = caudate nucleus; I = insula; L = lentiform nucleus; WL = window level; WW = window width.

brain region with that of contralateral normal-appearing white matter

(FIGURE 3-7).20 If the concerning brain region looks darker than the contralateral
normal-appearing white matter, the region has subacute-appearing changes.
These subacute-appearing regions are a marker of increased risk of hemorrhage
with thrombolysis.20 The presence of these regions does not mean that
thrombolysis is absolutely contraindicated, but they do mean that clinicians

294 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 KEY POINT

● On noncontrast CT, brain

regions that are darker
than the contralateral
normal-appearing white
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

matter are a marker of

increased risk of
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

hemorrhage with
thrombolysis. The presence
of these regions does not
mean that thrombolysis is
absolutely contraindicated,
but they do mean that
clinicians should proceed
with caution after weighing
the risks and benefits of
thrombolysis.
tbw+JMEi5ZgHD on 07/29/2023

FIGURE 3-7
Diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) MRI
changes with corresponding changes on noncontrast CT (NCCT). When a mismatch is seen
between DWI and FLAIR, noncontrast CT shows only subtle early ischemic changes (A, B, and
C). When no mismatch is seen between DWI and FLAIR, noncontrast CT shows
subacute-appearing changes (D, E, and F).
Reprinted with permission from Qazi E, et al, Med Res Arch.21 © 2016 KEI Journals.

should proceed with caution after weighing the risks and benefits of
thrombolysis. If subacute-appearing changes are present, the stroke symptom
onset time should be questioned again.21

HYPERDENSE ARTERY SIGN. A thrombus within the intracranial arterial tree is

often seen on the noncontrast CT as the hyperdense artery sign.22,23 A thrombus
that is rich in red blood cells is likely hyperdense. It is best to visualize the
hyperdense artery sign on thin slices (1.25-mm thickness or thinner).24,25 The
sign has high specificity (approximately 95%) but modest sensitivity
(approximately 50%) for detecting intraarterial thrombus when assessed later
by catheter angiography.26 False positives occur in patients with raised
hematocrit (eg, due to dehydration or smoking); false negatives occur when the
CT slices are thick (approximately 5 mm or thicker), and partial volume effects
reduce signal intensity within the region of interest or with a thrombus that is
more fibrin rich than red blood cell rich.

THE EYE DEVIATION SIGN. Finally, when rapidly assessing the noncontrast CT in
patients with acute ischemic stroke, it also helps to look at the patient’s eyes to see if
they are symmetrically deviated toward the suspected ischemic cerebral hemisphere
(CASE 3-1B). This radiologic sign does not have any value in isolation but is
useful to help focus attention on the cerebral hemisphere that is likely involved.

CONTINUUMJOURNAL.COM 295

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

CT Angiography
Acute stroke is a vascular disease caused either by leaking blood vessels
(hemorrhage) or a thrombus blocking the blood vessel (ischemic stroke).
Imaging the blood vessels, therefore, helps determine the etiology of an acute
stroke. In addition, vascular imaging helps with assessment of prognosis and in
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

making treatment decisions.

Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

The primary modality used to image blood vessels supplying the brain is the
CTA. It is best to acquire a head and neck CTA (aortic arch to vertex) to visualize
all extracranial and intracranial arteries supplying the brain (TABLE 3-3).27 It takes
about 2 more minutes to do head and neck CTA immediately after noncontrast
CT while the patient is lying on the CT gantry. The total radiation a patient receives
with a full stroke noncontrast CT/CTA head and neck protocol (7 millisieverts

CASE 3-1B The patient in CASE 3-1A had a noncontrast head CT. Rapid assessment of
tbw+JMEi5ZgHD on 07/29/2023

the CT (TABLE 3-1) revealed eye deviation toward the left, a hyperdense
left middle cerebral artery sign, early ischemic changes in the left
cerebral hemisphere (obscuration of the lentiform nucleus, the insular
ribbon sign, and loss of gray-white differentiation in the parietal cortex),
and an Alberta Stroke Program Early CT Score (ASPECTS) of 6, with no
evidence of any intracerebral, subarachnoid, subdural, or extradural
hemorrhage (FIGURE 3-8).

FIGURE 3-8
Noncontrast CT of the patient in CASE 3-1 confirmed eye deviation (A, arrow), and showed
a hyperdense vessel sign on the left (B, arrow) and early ischemic changes (loss of
grey-white matter differentiation) in the left M1, M2, insula (C, stars) and the left M6
region (D, star) on the Alberta Stroke Program Early CT Score (ASPECTS) template.

COMMENT The noncontrast CT (FIGURE 3-8) suggests the presence of a moderate-sized

ischemic stroke and potentially salvageable brain distal to a left M1
segment middle cerebral artery thrombus. This patient may benefit from
both IV thrombolysis and endovascular thrombectomy. Imaging modalities
such as CT angiography or CT perfusion (discussed later) can provide more
certainty about the intracranial artery occlusion site and presence or
absence of salvageable brain.

296 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 [mSv] to 9 mSv) compares favorably to a routine chest CT (7 mSv to 11 mSv) or an
abdomen CT (8 mSv to 12 mSv).27,28 Although the iodinated contrast agent used
for CTA results in a small risk of contrast-induced nephropathy (25% increase
in serum creatinine from baseline or a 0.5-mg/dL increase over 48 to 72 hours),
the risk is so minimal in most patients and the benefits of vascular imaging so
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

great that most stroke centers perform CTA immediately after noncontrast CT
without waiting for a serum creatinine.27 The only exception is in patients with
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

known severe renal insufficiency not on dialysis. If patients are on metformin, it

is recommended to withhold that drug for at least 48 hours after CTA.29
Some stroke centers modify imaging protocols for head and neck CTA to
include additional imaging of cerebral blood vessels over time. A commonly used
protocol is called multiphase CTA.28 This protocol acquires two additional head
CTA images taken approximately 8 and 16 seconds after the first acquisition.
Multiphase CTA does not use any additional contrast dose and only adds about
1 mSv additional radiation to the protocol. Multiphase CTA helps address the
poor CTA quality that happens when a delay occurs between acquisition of the
tbw+JMEi5ZgHD on 07/29/2023

scan and the arrival of contrast within the arteries of interest. In addition, by
providing images of contrast filling the arteries over time, multiphase CTA
provides useful information that helps with diagnosis, assessing prognosis, and
making treatment decisions, as discussed later in this article. Dynamic CTA is
another time-resolved CTA protocol that has similar advantages to multiphase
CTA but is derived from additional CTP imaging.30

HEMORRHAGE. As CTA helps image blood vessels, it is a useful tool to help

understand the etiology and pathophysiology of any intracranial hemorrhage.31
Pathologies such as intracranial aneurysms, arteriovenous malformations, dural
arteriovenous fistulas, and other vascular malformations can be detected using a
routine head CTA. The probability of finding these secondary causes of
intracranial hemorrhage ranges from 15% in adults to as high as 65% in patients
younger than 40 years of age.31 A head CTA is therefore recommended in patients
with intracranial hemorrhage who are young, have a lobar or infratentorial
location of hemorrhage, and do not have hypertension or impaired coagulation.

Imaging Features on CT Angiography to Rapidly Assess in Patients TABLE 3-3

Presenting With Acute Stroke

Intracranial Hemorrhage
◆ Abnormal-looking vessels
◆ Spot sign
Ischemia
◆ Presence and location of thrombus (intracranial or extracranial)
◆ Collateral status
◆ Any extracranial large artery pathology (plaque, dissection)
◆ Aortic disease
◆ Other pathologies (carotid web, vasculopathy)
◆ Arch and large vessel anatomy

CONTINUUMJOURNAL.COM 297

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

In patients with primary spontaneous intracerebral hemorrhage (likely due

to hypertension), CTA helps clinicians predict the risk of hemorrhage growth
over time. The spot sign is a serpiginous or linear contrast density located
within the parenchymal hemorrhage (FIGURE 3-9).32,33 The spot sign must not
be in anatomic continuity with any adjacent blood vessels. In addition, the
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

spot sign must be 120 Hounsfield units or more. When these criteria are not
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

met, the spot sign mimics, as these other contrast densities within or around
parenchymal hemorrhages are called, may be markers of other pathologies
such as choroidal calcifications, tumors, moyamoya disease, aneurysms, and
micro-arteriovenous malformations. Although two recent trials that used the spot
sign to tailor hemostatic therapy (recombinant factor VIIa) were not conclusive,34
the sign is useful as a marker of hemorrhage growth and can therefore be used to
tailor therapy such as blood pressure control or hemostasis and to help
determine prognosis.
With increasing use of multiphase or dynamic CTA in patients with acute
stroke, it has now become possible to assess the presence and size of the spot sign
tbw+JMEi5ZgHD on 07/29/2023

across the different time-resolved phases of a CTA.35 If the spot sign is seen in the
early arterial phase of a multiphase CTA, it is likely to be associated with greater
hemorrhage growth than when the sign is only seen in the venous phase of a
multiphase CTA (FIGURE 3-9). The rate of increase in size of the spot sign may
also be a marker of hemorrhage growth. More clinical trials are needed to
generate evidence on whether the detection and growth of spot signs may be
used to tailor therapy in patients with intracerebral hemorrhage.

ISCHEMIA. CTA of the head and neck is an essential tool in the management of
patients with acute ischemic stroke. It helps to detect thrombi within arteries
and their extent, collateral status beyond occlusive thrombus, and any other
associated pathologies. The tool also helps in determining the risk of recurrent
strokes and in planning acute endovascular treatment and surgical management
of carotid stenosis, as discussed below.

FIGURE 3-9
The spot sign, a marker of intracerebral hemorrhage growth, seen in the three phases (arrows)
of a multiphase CT angiogram.

298 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 PRESENCE AND LOCATION OF THROMBUS. In any patient with symptoms and signs KEY POINTS
of acute ischemic stroke, it is essential to know the vascular status of the large
● The primary modality
arteries (brachiocephalic [innominate] trunk, left carotid artery, and left used to image blood vessels
subclavian artery) arising from the aorta and their branches as they supply the supplying the brain is CT
brain. The head and neck CTA helps physicians managing acute stroke to angiography (CTA). It is best
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

identify the presence and location of any thrombus within this vasculature.27 to acquire a head and neck
CTA (aortic arch to vertex) to
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

The head and neck CTA, therefore, has the following important uses:
visualize all extracranial and
intracranial arteries
u Identifying thrombi in large proximal intracranial arteries (also called large vessel occlusions
supplying the brain.
[LVOs]). Patients with LVOs can benefit from acute endovascular treatment.
u Determining the location of thrombus within the intracranial arterial tree (eg, internal ● CTA is a useful tool to
carotid artery versus M1 segment of MCA versus more distal arteries) helps identify the help understand the
likelihood that a thrombus will respond to IV thrombolysis. Multiple studies have shown etiology of any intracranial
that an occlusive thrombus within the internal carotid artery is less likely to dissolve with hemorrhage and to identify
IV thrombolytics when compared to thrombus in more distal arteries.36,37 underlying pathologies, such
as intracranial aneurysms,
u Assessing thrombus burden (the extent of thrombus within the intracranial arterial tree) helps
arteriovenous
determine whether acute ischemic stroke treatment (endovascular therapy or IV thrombolysis)
tbw+JMEi5ZgHD on 07/29/2023

malformations, dural
is likely to be successful or not. A large
arteriovenous fistulas, and
thrombus burden is associated with reduced
any other vascular
efficacy of IV thrombolytics and potentially
malformations.
with longer endovascular treatment times.38
u In patients with minor strokes or transient ● On CTA, the spot sign is a
ischemic attacks, detecting thrombus within serpiginous or linear
the arterial vasculature supplying the brain contrast density located
identifies patients at risk of recurrent ischemic within the parenchymal
strokes. A composite brain CT/head and neck hemorrhage. The presence
CTA “at risk” metric (defined as presence of of a spot sign suggests
acute stroke on noncontrast CT, circle of Willis hemorrhage that is likely to
intracranial vessel occlusion or >50% stenosis, grow over time.
extracranial occlusion or >50% stenosis)
identifies patients at high risk of recurrent ● CTA is an essential tool in
ischemic stroke (negative predictive value of the management of patients
96% in data from the CATCH [CT and MRI in with acute ischemic stroke.
the Triage of TIA and minor Cerebrovascular It helps in detecting thrombi
Events to Identify High Risk Patients] study).39 within arteries and their
extent, collateral status
beyond occlusive thrombus,
NATURE OF THROMBUS. Thrombi within the and any other associated
intracranial arterial tree are either formed pathologies. The tool also
in situ or reach there as emboli. Most helps in determining the risk
FIGURE 3-10
intracranial arterial thrombi are embolic in of recurrent strokes and in
CT angiogram images that help
planning acute endovascular
nature; a thrombus can, however, form in characterize whether a thrombus is
treatment and surgical
situ even around emboli. A freshly formed porous. A, M1 segment middle cerebral
management of carotid
artery thrombus with residual flow
thrombus is red blood cell rich, but over stenosis.
grade 0: no contrast permeation of
time, thrombi change to become more thrombus (arrowhead). B, Thrombus in
fibrin rich.40–42 Red blood cell–rich thrombi the same location with residual flow
are likely more permeable to blood flow grade 1: contrast permeating diffusely
and, therefore, to thrombolytics.43 through thrombus (arrowheads). C,
Thrombus in the same location with
CTA can be used to detect intracranial residual flow grade 2: tiny hairline
thrombus that is permeable to IV contrast lumen or streak of well-defined
and therefore to blood flow. The arterial contrast within the thrombus extending
silhouette on the CTA is used to identify either through its entire length or
through part of thrombus (arrowhead).
contrast permeation within the site of Thrombi with residual flow are more
intracranial thrombus. This construct is called likely to lyse with IV recombinant tissue
residual flow on CTA (FIGURE 3-10). Residual plasminogen activator (rtPA).

CONTINUUMJOURNAL.COM 299

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

flow seen on CTA is graded: (1) grade 0—no contrast permeation of the
thrombus, (2) grade 1—contrast permeating diffusely through thrombus, or (3)
grade 2—tiny hairline lumen or a streak of well-defined contrast within the
thrombus extending either through its entire length or through part of the
thrombus. Intracranial thrombi with grade 1 or 2 residual flow on CTA are more
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

likely to dissolve early with IV recombinant tissue plasminogen activator (rtPA)

than thrombi with no residual flow.36
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

Another imaging construct that helps determine permeable thrombi on CTA

is thrombus attenuation increase. Thrombus attenuation increase involves
subtracting the mean Hounsfield units of the thrombus on noncontrast CT from
the mean Hounsfield units of the thrombus on CTA. A difference of 10 units or
more is suggestive of a permeable thrombus that is more amenable to lysis with
thrombolytics.44 In the author’s experience, however, a maximal thrombus
attenuation unit in the region of the thrombus on CTA greater than 90
Hounsfield units is suggestive of a permeable thrombus.
tbw+JMEi5ZgHD on 07/29/2023

COLLATERAL STATUS. Collaterals are direct anastomoses between cerebral

blood vessels that allow rerouting of blood flow when direct blood supply is
blocked to brain tissue.45 Since collaterals provide blood flow to ischemic brain
tissue distal to blocked arteries, assessing collateral status is a good method to

FIGURE 3-11
Multiphase CT angiogram assessment of collateral status in the symptomatic brain in three
phases. A, A one-phase delay is seen in filling of arterial contrast between the left
(symptomatic) side and right side. B, Impaired washout of contrast is seen in the third phase
on the left (symptomatic) side when compared to the right side. C, A reduced number of
vessels is seen on the left (symptomatic) side when compared to the right side.
Reprinted with permission from d’Esterre CD, et al, Stroke.46 © 2017 Wolters Kluwer Health.

300 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 assess the degree of blood flow to ischemic brain tissue, and multiphase CTA is KEY POINT
an excellent tool to assess collateral status.28 On the three time-resolved phases
● Multiphase CTA is an
of the multiphase CTA, arteries distal to the blocked artery are assessed for excellent tool to assess
extent of arterial contrast, delay in filling of contrast, and impaired washout collateral status. On the
of contrast when compared to arteries on the contralateral side three time-resolved phases
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

(FIGURE 3-11 and FIGURE 3-12). of the multiphase CTA,

arteries distal to the blocked
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

Areas of the ischemic brain with minimal or no vessels are indicative of severe artery are assessed for
ischemia resulting in irreversibly injured brain; ischemic brain with a one-phase extent of arterial contrast,
delay in contrast filling with impaired washout of that contrast in the third phase is delay in filling of contrast,
indicative of severely ischemic brain that can be saved with quick reperfusion; and impaired washout of
contrast when compared
ischemic brain with one-phase delay in contrast filling with washout of contrast to arteries on the
in the third phase is indicative of mildly ischemic brain (FIGURE 3-12, contralateral side.
CASE 3-1C). These constructs on multiphase CTA have correlates on CTP
imaging.46 Recent advances in image processing have resulted in composite
time-resolved images of collaterals that combine information from all phases
of a multiphase CTA (FIGURE 3-14).
tbw+JMEi5ZgHD on 07/29/2023

FIGURE 3-12
Multiphase CT angiogram assessment of collaterals in three cortical regions (M4, M5, and
M6). The left brain is symptomatic. In the M6 region, minimal vessels are seen in all
three phases, suggestive of very severe ischemia and potentially nonsalvageable brain
(confirmed on 24-hour follow-up noncontrast CT [NCCT]) even with reperfusion. In the M5
region, a one-phase delay of arterial filling is seen with impaired washout of contrast. This is
suggestive of moderate ischemia and brain that may be salvageable with reperfusion. In the
M4 region, a one-phase delay is seen but with early washout of contrast within arteries when
compared to the corresponding brain region on the right side, suggesting mild ischemia. The
corresponding CT perfusion transit time (Tmax) and cerebral blood flow (CBF) parameters
(bottom panel) predict tissue fate similarly.

CONTINUUMJOURNAL.COM 301

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

IDENTIFYING OTHER PATHOLOGIES. The CTA also helps identify other

relevant pathologies in patients presenting with acute stroke, including
disease of the aorta (thrombus, plaque, and dissection [FIGURE 3-15]),
extracranial atherosclerotic disease, arterial dissections, carotid webs
(FIGURE 3-16), and intracranial arterial narrowing (focal vasculopathy
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

or vasculitis).
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

PLANNING NEUROINTERVENTION. Head and neck CTA is an important tool in the

armamentarium of interventional neurologists in planning acute endovascular
thrombectomy. Assessment of aortic arch and large artery anatomy helps in
choosing the type of catheter to be used during the procedure. In addition, the
location and extent of thrombus within the arterial tree also helps determine the
device type and profile used for mechanical thrombectomy.
tbw+JMEi5ZgHD on 07/29/2023

CASE 3-1C The patient in CASE 3-1A had a multiphase head and neck CT angiogram
(CTA). Rapid assessment of the CTA (TABLE 3-3) revealed an intracranial
thrombus in the left middle cerebral artery, moderate collaterals (one-
phase delay with impaired washout of contrast within arteries distal to
the thrombus when compared to the right side) (FIGURE 3-13), and a plaque
at the left carotid artery bifurcation with stenosis of approximately 70%
(tandem occlusion).

FIGURE 3-13
Multiphase CT angiography from the patient in CASE 3-1. Phase 1 (A), phase 2 (B), and
phase 3 (C) show intermediate collaterals with a one-phase delay and reduced extent
(arrowheads).

COMMENT Patients with large vessel occlusions (LVOs) are ideal targets for
endovascular therapy when the brain beyond the occlusion is supplied by
moderate to good collaterals. The multiphase CTA of the head and neck
helped identify the presence of an LVO (left M1 segment middle cerebral
artery occlusion) with moderate collaterals distal to the occlusion.

302 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 CT Perfusion KEY POINTS
Like CTA, CTP is a CT scan obtained after
● Head and neck CTA is an
the injection of IV contrast. Like the important tool in planning
multiphase CTA, the technique of CTP acute endovascular
involves acquiring multiple scans of the thrombectomy. Assessment
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

brain over time; summating these of the aortic arch and large
artery anatomy helps in
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

time-resolved images of contrast filling in

choosing the type of
and washing out from brain using catheter to be used during
mathematical formulas (algorithms such the procedure. In addition,
as deconvolution); and generating the location and extent of
thrombus within the arterial
estimates of cerebral blood flow, blood
tree also helps determine
volume, and transit time within brain the device type and profile
tissue. Unlike multiphase CTA, which only used for mechanical
has three time-resolved images, CTP thrombectomy.
involves acquiring many such images. This
● CT perfusion (CTP)
results in a higher radiation dose than in
tbw+JMEi5ZgHD on 07/29/2023

involves acquiring multiple

multiphase CTA.1,28 Depending on the FIGURE 3-14 scans of the brain over time;
hospital’s scanner, CTP may only acquire Composite time-resolved map of summating these time-
collateral filling generated using resolved images of contrast
images from part of the brain. Hospital multiphase CT angiography. The filling in and washing out
scanning protocols are also different; if occlusion is in the right proximal M1 from brain using
CTP images are acquired over 40 seconds segment of the middle cerebral artery mathematical formulas; and
or less, blood volume is underestimated (arrow). The algorithm classifies all generating estimates of
voxels into three categories (red, green,
(called truncation artifact).47,48 Moreover, cerebral blood flow, blood
or blue) based on whether maximum volume, and transit time
the algorithms that are used to make contrast enhancement appears before within brain tissue.
calculations on CTP images are varied; or after an automatically identified
physicians should therefore understand venous window for that patient. ● Current CTP techniques
Arteries appearing red, therefore, show
which algorithm is used at their hospital are prediction tools. They
peak contrast early, arteries appearing help predict the probability
and the output characteristics of that green show peak contrast at a time of brain tissue being dead or
particular algorithm. CTP may be used in similar to when contrast density peaks alive by estimating the
acute stroke to predict ischemic infarct in veins on the normal side, and arteries degree of blood flow within
and salvageable brain and to predict the appearing blue show peak contrast that tissue.
later than when contrast density peaks
risk of intracerebral hemorrhage after in veins on the normal side.
acute stroke treatment.

PREDICTING ISCHEMIC INFARCT AND SALVAGEABLE BRAIN. Current CTP

techniques are prediction tools. They help predict the probability of brain
tissue being dead or alive by estimating the degree of blood flow within that
tissue.49–51 Brain that has a reasonably high probability of being dead even after
reperfusion is called the ischemic core. Brain that is not functioning but is
salvageable after reperfusion is called the penumbra. Since CTP is a snapshot in
time of blood flow and not of brain tissue status and since brain tissue does not
infarct instantaneously but over time, no one blood flow value corresponds
precisely to dead brain. In patients who present early and in whom acute stroke
treatment is likely to result in early reperfusion, blood flow values that predict
dead tissue are likely to be far lower than in patients who present late or are
treated late (FIGURE 3-17).51 If brain blood flow changes over time because of
hemodynamic fluctuations (eg, blood pressure changes), the value of the
snapshot CTP image in predicting brain tissue fate is reduced. This is unlike the
value of hypoattenuation on noncontrast CT, which is a marker of
brain infarction.

CONTINUUMJOURNAL.COM 303

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0
tbw+JMEi5ZgHD on 07/29/2023

FIGURE 3-15
Imaging of a 59-year-old man with left hemiparesis after a minor motor vehicle accident. CT
angiography shows dissection of the ascending (A, arrow), transverse (B, arrow), and
descending aorta with extension to cervical carotid arteries (C, arrow) and occlusion of
right common carotid artery (D, arrow). Proximal intracranial arteries are patent (E), and a
small area of early ischemic changes is seen in the right frontal region (F, arrow).

For clinical purposes, therefore, it is important to understand that CTP output

maps are probability maps and do not reflect brain tissue status. The following
steps are suggested in interpreting CTP maps:

u Understand the type of CTP acquisition used at your hospital (ie, extent of brain coverage,
duration of scan acquisition). This will help in determining how much of the brain is imaged
and, in particular, whether estimates for cerebral blood volume are reliable.
u Understand the type of algorithms that are used at your hospital and the output
characteristics these algorithms generate. Different commercial vendors and automated
software provide different output characteristics.
u Recognize that, if the patient moves his or her head a lot during CTP acquisition (scanning
takes at least 40 to 60 seconds), image interpretation will be difficult and output
characteristics unreliable, even with the most advanced motion correction software.
u Always look at the early ischemic changes on noncontrast CT as those changes better
reflect brain tissue status than CTP, which only predicts that same tissue status using a
measure such as blood flow.
u Recognize that CTP predictions are not perfect and may fail.

PREDICTING RISK OF INTRACEREBRAL HEMORRHAGE AFTER ACUTE STROKE

52,53
TREATMENT. Brain tissue with very low blood flow is likely to infarct early.

304 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 Severe ischemia is also likely to KEY POINTS
cause endothelial injury to blood
● Brain tissue with very low
vessels supplying ischemic brain, blood flow on CTP is likely
thus making these vessels leaky. to infarct early. By detecting
By detecting regions of very low regions of very low blood
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

blood flow (or blood volume) or flow (or blood volume) or

regions of brain with
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

regions of brain with increased increased blood-brain

blood-brain barrier permeability, barrier permeability, CTP
CTP can help predict brain can help predict brain
regions with increased risk of regions with increased risk
of hemorrhage after acute
hemorrhage after acute stroke stroke treatment.
treatment (eg, thrombolysis).20,52,53
Subacute-appearing changes on ● Subacute-appearing
noncontrast CT or regions of the changes on noncontrast CT
or regions of the brain
brain with changes seen on
with changes seen on
diffusion-weighted images
tbw+JMEi5ZgHD on 07/29/2023

diffusion-weighted images
(DWI) and fluid-attenuated (DWI) and fluid-attenuated
FIGURE 3-16
inversion recovery (FLAIR) inversion recovery (FLAIR)
Sagittal reconstruction of a neck CT angiogram images (no DWI-FLAIR
images (no DWI-FLAIR showing a web (arrow) at the carotid bulb. mismatch) may also help
mismatch; FIGURE 3-7) may also clinicians predict the risk of
help clinicians predict the risk of hemorrhage with acute
hemorrhage with acute treatment similarly.20,21,54 treatment.

PATIENT SELECTION FOR ACUTE STROKE THERAPY

The principles of clinical decision making in acute stroke include quick history
taking and examination followed by imaging that is readily available and easy to
interpret, leading to quick treatment decisions (CASE 3-1D). This workflow is the
norm for most clinical situations. Even in patients who present many hours after
acute stroke onset, were last known well many hours earlier, or woke up with
stroke symptoms, the same principles of decision making apply. The goal is to

FIGURE 3-17
Time-based CT perfusion thresholds that help predict follow-up infarction.
rel CBF = relative cerebral blood flow; Tmax = a contrast transit time parameter.
Reprinted with permission from d’Esterre CD, et al, Stroke.50 © 2015 Wolters Kluwer Health.

CONTINUUMJOURNAL.COM 305

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

assess imaging to determine the type of stroke, assess markers of prognosis and
treatment risk such as the extent and severity of ischemia or hemorrhage,
identify treatment targets such as the presence of a large vessel occlusion in
ischemic stroke or the spot sign in patients with hemorrhage, and make decisions.
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

CONCLUSION
Time is of the essence when making decisions when treating patients with acute
stroke. Given its ease of use, availability, and speed of acquisition, physicians

CASE 3-1D To summarize, the patient in

CASE 3-1A had presented to the
emergency department with
severe symptoms suggestive
tbw+JMEi5ZgHD on 07/29/2023

of acute stroke 3 hours from

symptom onset. A quick
history did not reveal any
contraindications to acute
stroke treatment. Noncontrast
CT showed early ischemic
changes, with an Alberta
Stroke Program Early CT
Score (ASPECTS) of 6.
Multiphase CT angiography
(CTA) showed a left middle
cerebral artery M1 segment
thrombus, no residual flow,
and moderate collaterals with
tandem 70% stenosis of the
carotid artery at the neck.
The patient received IV FIGURE 3-18
recombinant tissue CT angiogram of the patient in CASE 3-1. The initial CT
angiogram (A) confirmed the suspected M1
plasminogen activator (rtPA) segment middle cerebral artery occlusion (arrow).
at a dose of 0.9 mg/kg body After thrombus retrieval with a stent retriever,
weight (10% as bolus and the complete reperfusion was restored (B).
remaining as infusion over
60 minutes) and was
immediately taken to the angiography suite for acute endovascular
therapy. The first angiographic acquisition confirmed the presence of the
left M1 segment middle cerebral artery occlusion (FIGURE 3-18A).
Endovascular thrombectomy with a stent retriever resulted in complete
reperfusion (FIGURE 3-18B). He showed clinical improvement after
reperfusion was achieved.

COMMENT This patient’s case is an example of the principles of clinical decision

making in acute stroke.

306 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 should use CT as the primary imaging tool in these patients. Noncontrast head
CT and CTA of the head and neck are essential modalities in patients with acute
stroke, helping to diagnose stroke, determine prognosis, and identify patients
who might benefit from acute stroke treatment. Techniques such as multiphase
CTA or CTP provide additional assistance in identifying salvageable brain and
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

assessing the risk of treatment. Acute stroke medicine is an art and science that
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

involves rapidly assessing relevant clinical and imaging information and making
clinical decisions in a timely manner.

REFERENCES

1 Menon BK, Goyal M. Imaging paradigms in acute 13 Dzialowski I, Klotz E, Goericke S, et al. Ischemic
ischemic stroke: a pragmatic evidence-based brain tissue water content: CT monitoring during
approach. Radiology 2015;277(1):7–12. doi:10.1148/ middle cerebral artery occlusion and reperfusion
radiol.2015151030. in rats. Radiology 2007;243(3):720–726. doi:10.
1148/radiol.2432060137.
2 Saver JL. Time is brain—quantified. Stroke 2006;
tbw+JMEi5ZgHD on 07/29/2023

37(1):263–266. doi:10.1161/01.STR.0000196957. 14 Puetz V, Dzialowski I, Hill MD, Demchuk AM. The

55928.ab. Alberta Stroke Program Early CT Score in clinical
practice: what have we learned? Int J Stroke
3 Lyden P, Raman R, Liu L, et al. National Institutes
2009;4(5):354–364. doi:10.1111/j.1747-4949.2009.
of Health Stroke Scale certification is reliable
00337.x.
across multiple venues. Stroke 2009;40(1):
2507–2511. doi:10.1161/STROKEAHA.108.532069. 15 Kent DM, Hill MD, Ruthazer R, et al. “Clinical-CT
mismatch” and the response to systemic
4 von Kummer R, Dzialowski I. MRI versus CT in
thrombolytic therapy in acute ischemic stroke.
acute stroke. Lancet 2007;369(9570):1341–1342.
Stroke 2005;36(8):1695–1699. doi:10.1161/01.
doi:10.1016/S0140-6736(07)60621-7.
STR.0000173397.31469.4b.
5 Ciccone A, Sterzi R, Munari L, students assessing
16 Kalafut MA, Schriger DL, Saver JL, Starkman S.
the paper. MRI versus CT in acute stroke.
Detection of early CT signs of >1/3 middle
Lancet 2007;369(21):1342–1343. doi:10.1016/
cerebral artery infarctions: interrater reliability
S0140-6736(07)60624-2.
and sensitivity of CT interpretation by physicians
6 Al-Shahi Salman R, Labovitz DL, Stapf C. involved in acute stroke care. Stroke 2000;31(7):
Spontaneous intracerebral haemorrhage. BMJ 1667–1671. doi:10.1161/01.str.31.7.1667.
2009;339:b2586. doi:10.1136/bmj.b2586.
17 Barber PA, Demchuk AM, Zhang J, Buchan AM.
7 Kothari RU, Brott T, Broderick JP, et al. The ABCs Validity and reliability of a quantitative computed
of measuring intracerebral hemorrhage volumes. tomography score in predicting outcome of
Stroke 1996;27(8):1304–1305. doi:10.1161/01. hyperacute stroke before thrombolytic therapy.
str.27.8.1304. ASPECTS Study Group Alberta Stroke Programme
Early CT Score Lancet 2000;355(9216):1670–1674.
8 Broderick JP, Brott TG, Duldner JE, et al. Volume
doi:10.1016/s0140-6736(00)02237-6.
of intracerebral hemorrhage. A powerful and
easy-to-use predictor of 30-day mortality. Stroke 18 Menon BK, Puetz V, Kochar P, Demchuk AM.
1993;24(7):987–993. doi:10.1161/01.str.24.7.987. ASPECTS and other neuroimaging scores in the
triage and prediction of outcome in acute stroke
9 Broderick JP, Brott TG, Grotta JC. Intracerebral
patients. Neuroimaging Clin N Am 2011;21(2):
hemorrhage volume measurement. Stroke 1994;
407–423, xii. doi:10.1016/j.nic.2011.01.007.
25(5):1081. doi:10.1161/str.25.5.1081b.
19 Wilson AT, Dey S, Evans JW, et al. Minds treating
10 Sporns PB, Kemmling A, Schwake M, et al. Triage
brains: understanding the interpretation of non-
of 5 noncontrast computed tomography markers
contrast CT ASPECTS in acute ischemic stroke.
and spot sign for outcome prediction after
Expert Rev Cardiovasc Ther 2018;16(2):143–153.
intracerebral hemorrhage. Stroke 2018;49(10):
doi:10.1080/14779072.2018.1421069.
2317–2322. doi:10.1161/STROKEAHA.118.021625.
20 Batchelor C, Pordeli P, d’Esterre CD, et al. Use of
11 Baron JC. Perfusion thresholds in human cerebral
noncontrast computed tomography and
ischemia: historical perspective and therapeutic
computed tomographic perfusion in predicting
implications. Cerebrovasc Dis 2001;11(suppl 1):
intracerebral hemorrhage after intravenous
2–8. doi:10.1159/000049119.
alteplase therapy. Stroke 2017;48(6):1548–1553.
12 Dzialowski I, Weber J, Doerfler A, et al. Brain doi:10.1161/STROKEAHA.117.016616.
tissue water uptake after middle cerebral artery
occlusion assessed with CT. J Neuroimaging 2004;
14(1):42–48. doi:10.1111/j.1552-6569.2004.tb00214.x.

CONTINUUMJOURNAL.COM 307

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 NEUROIMAGING IN ACUTE STROKE

21 Qazi E, Al-Ajlan FS, Mahajan A, et al. Non-contrast 33 Demchuk AM, Dowlatshahi D, Rodriguez-Luna D,
CT in place of MRI mismatch in the imaging triage et al. Prediction of haematoma growth and
of acute ischemic stroke patients. Med Res outcome in patients with intracerebral
Arch 2016;4(6). haemorrhage using the CT-angiography spot sign
(PREDICT): a prospective observational study.
22 Schuierer G, Huk W. The unilateral hyperdense
Lancet Neurol 2012;11(4):307–314. doi:10.1016/
middle cerebral artery: an early CT-sign of
S1474-4422(12)70038-8.
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

embolism or thrombosis. Neuroradiology 1988;

30(2):120–122. doi:10.1007/BF00395612. 34 Gladstone DJ, Aviv RI, Demchuk AM, et al. Effect
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

of recombinant activated coagulation factor VII

23 Tomsick TA, Brott TG, Chambers AA, et al.
on hemorrhage expansion among patients
Hyperdense middle cerebral artery sign on CT:
with spot sign-positive acute intracerebral
efficacy in detecting middle cerebral artery
hemorrhage: the SPOTLIGHT and STOP-IT
thrombosis. AJNR Am J Neuroradiol 1990;11:
randomized clinical trials. JAMA Neurol 2019.
473–477. doi:10.1007/bf00395612.
doi:10.1001/jamaneurol.2019.2636.
24 Riedel CH, Jensen U, Rohr A, et al. Assessment
35 Chakraborty S, Alhazzaa M, Wasserman JK, et al.
of thrombus in acute middle cerebral artery
Dynamic characterization of the CT angiographic
occlusion using thin-slice nonenhanced
‘spot sign’. PLoS One 2014;9:e90431. doi:10.1371/
Computed Tomography reconstructions. Stroke
journal.pone.0090431.
2010;41(8):1659–1664. doi:10.1161/STROKEAHA.
110.580662. 36 Menon BK, Al-Ajlan FS, Najm M, et al. Association
of clinical, imaging, and thrombus characteristics
25 Riedel CH, Zoubie J, Ulmer S, Gierthmuehlen J,
tbw+JMEi5ZgHD on 07/29/2023

with recanalization of visible intracranial

Jansen O. Thin-slice reconstructions of
occlusion in patients with acute ischemic stroke.
nonenhanced CT images allow for detection of
JAMA 2018;320(10):1017–1026. doi:10.1001/
thrombus in acute stroke. Stroke 2012;43(9):
jama.2018.12498.
2319–2323. doi:10.1161/STROKEAHA.112.649921.
37 Mishra SM, Dykeman J, Sajobi TT, et al. Early
26 Mair G, Boyd EV, Chappell FM, et al. Sensitivity
reperfusion rates with IV tPA are determined by
and specificity of the hyperdense artery sign for
CTA clot characteristics. AJNR Am J Neuroradiol
arterial obstruction in acute ischemic stroke.
2014;35(12):2265–2272. doi:10.3174/ajnr.A4048.
Stroke 2015;46(1):102–107. doi:10.1161/
STROKEAHA.114.007036. 38 Puetz V, Dzialowski I, Hill MD, et al. Intracranial
thrombus extent predicts clinical outcome, final
27 Demchuk AM, Menon BK, Goyal M. Comparing
infarct size and hemorrhagic transformation in
vessel imaging: noncontrast computed
ischemic stroke: the clot burden score. Int J
tomography/computed tomographic
Stroke 2008;3(4):230–236. doi:10.1111/j.1747-4949.
angiography should be the new minimum
2008.00221.x.
standard in acute disabling stroke. Stroke 2016;
47(1):273–281. doi:10.1161/STROKEAHA.115.009171. 39 Coutts SB, Modi J, Patel SK, et al. CT/CT
angiography and MRI findings predict recurrent
28 Menon BK, d'Esterre CD, Qazi EM, et al.
stroke after transient ischemic attack and minor
Multiphase CT angiography: a new tool for the
stroke: results of the prospective CATCH study.
imaging triage of patients with acute ischemic
Stroke 2012;43(4):1013–1017. doi:10.1161/
stroke. Radiology 2015;275(2):510–520. doi:10.
STROKEAHA.111.637421.
1148/radiol.15142256.
40 Almekhlafi MA, Hu WY, Hill MD, Auer RN.
29 Baerlocher MO, Asch M, Myers A. Five things
Calcification and endothelialization of thrombi in
to know about...metformin and intravenous
acute stroke. Ann Neurol 2008;64(3):344–348.
contrast. CMAJ 2013;185(1):E78. doi:10.1503/
doi:10.1002/ana.21404.
cmaj.090550.
41 Marder VJ, Chute DJ, Starkman S, et al. Analysis
30 Menon BK, O'Brien B, Bivard A, et al. Assessment
of thrombi retrieved from cerebral arteries of
of leptomeningeal collaterals using dynamic CT
patients with acute ischemic stroke. Stroke 2006;
angiography in patients with acute ischemic
37(8):2086–2093. doi:10.1161/01.STR.0000230307.
stroke. J Cereb Blood Flow Metab 2013;33(3):
03438.94.
365–371. doi:10.1038/jcbfm.2012.171.
42 Brinjikji W, Duffy S, Burrows A, et al. Correlation of
31 Menon BK, Demchuk AM. Computed tomography
imaging and histopathology of thrombi in acute
angiography in the assessment of Patients with
ischemic stroke with etiology and outcome: a
stroke/TIA. Neurohospitalist 2011;1(4):187–199.
systematic review. J Neurointerv Surg 2017;9(6):
doi:10.1177/1941874411418523.
529–534. doi:10.1136/neurintsurg-2016-012391.
32 Thompson AL, Kosior JC, Gladstone DJ, et al.
43 Qazi EM, Sohn SI, Mishra S, et al. Thrombus
Defining the CT angiography ‘spot sign’ in
characteristics are related to collaterals and
primary intracerebral hemorrhage. Can J Neurol
angioarchitecture in acute stroke. Can J Neurol
Sci 2009;36(4):456–461. doi:10.1017/
Sci 2015;42(6):381–388. doi:10.1017/cjn.2015.291.
s0317167100007782.

308 APRIL 2020

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 44 Santos EM, Marquering HA, den Blanken MD, 50 d'Esterre CD, Boesen ME, Ahn SH, et al.
et al. Thrombus permeability is associated with Time-dependent computed tomographic
improved functional outcome and recanalization perfusion thresholds for patients with acute
in patients with ischemic stroke. Stroke 2016; ischemic stroke. Stroke 2015;46(12):3390–3397.
47(3):732–741. doi:10.1161/STROKEAHA.115.011187. doi:10.1161/STROKEAHA.115.009250.
45 Liebeskind DS. Collateral circulation. Stroke 51 Qiu W, Kuang H, Lee TY, et al. Confirmatory study
w3rlhcBKDSfJ2MrShoDNDnNFLYP4JCBgqQRwPS91wggrUrdwNAFueiPnB6S4p5a0pbvZocKn+cJ6A9e/GPJ3+PNxolCJkyhu3y8D

2003;34(9):2279–2284. doi:10.1161/01.STR. of time-dependent computed tomographic

0000086465.41263.06. perfusion thresholds for use in acute ischemic
Downloaded from http://journals.lww.com/continuum by NAI6iVqAdXD6eW7b5T7b7VEtuzqzHf92x8hJHD+v+Ga7NT0

stroke. Stroke 2019:STROKEAHA119026281. doi:10.

46 d'Esterre CD, Trivedi A, Pordeli P, et al.
1161/STROKEAHA.119.026281.
Regional comparison of multiphase computed
tomographic angiography and computed 52 Mishra NK, Christensen S, Wouters A, et al.
tomographic perfusion for prediction of tissue Reperfusion of very low cerebral blood volume
fate in ischemic stroke. Stroke 2017;48(4): lesion predicts parenchymal hematoma after
939–945. doi:10.1161/STROKEAHA.116.015969. endovascular therapy. Stroke 2015;46(5):
1245–1249. doi:10.1161/STROKEAHA.114.008171.
47 Copen WA, Deipolyi AR, Schaefer PW, et al.
Exposing hidden truncation-related errors in acute 53 Campbell BC, Christensen S, Butcher KS, et al.
stroke perfusion imaging. AJNR Am J Neuroradiol Regional very low cerebral blood volume
2015;36(4):638–645. doi:10.3174/ajnr.A4186. predicts hemorrhagic transformation better
than diffusion-weighted imaging volume and
48 Kasasbeh AS, Christensen S, Straka M, et al.
thresholded apparent diffusion coefficient in
Optimal computed tomographic perfusion scan
tbw+JMEi5ZgHD on 07/29/2023

acute ischemic stroke. Stroke 2010;41(1):82–88.

duration for assessment of acute stroke lesion
doi:10.1161/STROKEAHA.109.562116.
volumes. Stroke 2016;47(12):2966–2971. doi:10.
1161/strokeaha.116.014177. 54 Coutts SB, Menon BK. Late thrombolysis for
stroke works, but how do we do it? Lancet 2019;
49 Campbell BCV, Majoie CBLM, Albers GW, et al.
394(10193):97–98. doi:10.1016/S0140-6736(19)
Penumbral imaging and functional outcome in
31095-5.
patients with anterior circulation ischaemic
stroke treated with endovascular thrombectomy
versus medical therapy: a meta-analysis of
individual patient-level data. Lancet Neurol 2019;
18(1):46–55. doi:10.1016/S1474-4422(18)30314-4.

CONTINUUMJOURNAL.COM 309

Copyright © American Academy of Neurology. Unauthorized reproduction of this article is prohibited.