Position Paper

Neuroimaging standards for research into small vessel

disease and its contribution to ageing and
neurodegeneration
Joanna M Wardlaw, Eric E Smith, Geert J Biessels, Charlotte Cordonnier, Franz Fazekas, Richard Frayne, Richard I Lindley, John T O’Brien,
Frederik Barkhof, Oscar R Benavente, Sandra E Black, Carol Brayne, Monique Breteler, Hugues Chabriat, Charles DeCarli, Frank-Erik de Leeuw,
Fergus Doubal, Marco Duering, Nick C Fox, Steven Greenberg, Vladimir Hachinski, Ingo Kilimann, Vincent Mok, Robert van Oostenbrugge,
Leonardo Pantoni, Oliver Speck, Blossom C M Stephan, Stefan Teipel, Anand Viswanathan, David Werring, Christopher Chen, Colin Smith,
Mark van Buchem, Bo Norrving, Philip B Gorelick, Martin Dichgans; STandards for ReportIng Vascular changes on nEuroimaging (STRIVE v1)

Lancet Neurol 2013; 12: 822–38 Cerebral small vessel disease (SVD) is a common accompaniment of ageing. Features seen on neuroimaging include
This online publication has recent small subcortical infarcts, lacunes, white matter hyperintensities, perivascular spaces, microbleeds, and brain
been corrected. atrophy. SVD can present as a stroke or cognitive decline, or can have few or no symptoms. SVD frequently coexists
The corrected version first
with neurodegenerative disease, and can exacerbate cognitive deficits, physical disabilities, and other symptoms of
appeared at thelancet.com/
neurology on July 22, 2013 neurodegeneration. Terminology and definitions for imaging the features of SVD vary widely, which is also true for
Neuroimaging Sciences protocols for image acquisition and image analysis. This lack of consistency hampers progress in identifying the
(Prof J M Wardlaw MD), Centre contribution of SVD to the pathophysiology and clinical features of common neurodegenerative diseases. We are an
for Clinical Brain Sciences international working group from the Centres of Excellence in Neurodegeneration. We completed a structured
(Prof J M Wardlaw, C Smith MD),
process to develop definitions and imaging standards for markers and consequences of SVD. We aimed to achieve the
and Centre for Cognitive
Ageing and Cognitive following: first, to provide a common advisory about terms and definitions for features visible on MRI; second, to
Epidemiology suggest minimum standards for image acquisition and analysis; third, to agree on standards for scientific reporting
(Prof J M Wardlaw), Brain of changes related to SVD on neuroimaging; and fourth, to review emerging imaging methods for detection and
Research Imaging Centre
quantification of preclinical manifestations of SVD. Our findings and recommendations apply to research studies,
(F Doubal PhD), University of
Edinburgh, Edinburgh, UK; and can be used in the clinical setting to standardise image interpretation, acquisition, and reporting. This Position
Departments of Clinical Paper summarises the main outcomes of this international effort to provide the STandards for ReportIng Vascular
Neurosciences and Radiology, changes on nEuroimaging (STRIVE).
Hotchkiss Brain Institute,
University of Calgary and
Seaman Family MR Research Introduction Interpretation of data from many reports is hampered
Centre, Calgary, AL, Canada Neurodegenerative diseases such as Alzheimer’s disease by the variable consequences of acute SVD and related
(E E Smith MD, commonly coexist with cerebrovascular disease in older lesions and the convergence of lesions with different
Prof R Frayne PhD); Department
of Neurology, Rudolf Magnus
people. Cerebral small vessel disease (SVD) is the most causes but similar late appearances on MRI (figure 1).
Institute of Neuroscience, UMC common vascular cause of dementia, a major contributor Use of more standard terminology, definitions, and
Utrecht, Utrecht, Netherlands to mixed dementia, and the cause of about a fifth of all methods for image acquisition and analysis across
(G J Biessels MD); Department strokes worldwide.1,2 Alzheimer’s disease and SVD share research centres would remove a major barrier to
of Neurology, Lille University
Hospital, Lille, France
risk factors3,4 and both lead to cognitive decline and progress. Furthermore, such standardisation could be
(Prof C Cordonnier MD); dementia;5–7 the clinical differentiation of Alzheimer’s used in clinical practice to improve diagnosis and better
Department of Neurology, disease from vascular cognitive impairment or vascular understand the cause of cognitive impairment in
Medical University of Graz, dementia is increasingly recognised to be blurred.8 elderly patients.
Graz, Austria
(Prof F Fazekas MD); University
Signs of SVD on conventional MRI include recent Neuroimaging consensus standards for classification
of Sydney and George Institute small subcortical infarcts, white matter magnetic of SVD were first proposed by the US National Institute
for Global Health, Westmead resonance (MR) hyperintensities, lacunes, prominent of Neurological Disorders and Stroke and the Canadian
Hospital, University of Sydney, perivascular spaces, cerebral microbleeds, and atrophy.2 Stroke Network as part of the development of standards
Sydney, NSW, Australia
(Prof R I Lindley MD);
However, the terms for and definitions of these lesions for research on vascular cognitive impairment.11
Department of Psychiatry, have varied substantially between studies.9,10 For Subsequently, a scientific statement from the American
University of Cambridge, example, our systematic review identified 1144 instances Heart Association incorporated neuroimaging evidence
Addenbrooke’s Hospital,
of 50 different terms used to describe white matter for SVD or stroke as part of the criteria for probable
Cambridge, UK
(Prof J T O’Brien DM); hyperintensities in 940 papers; in some cases, two vascular mild cognitive impairment and dementia, and
Department of Radiology and different terms were used in the same paper (table 1; included a class 2 recommendation for neuroimaging
Nuclear Medicine, VU appendix). This amount of variation inhibits cross- as part of the clinical investigation of vascular cognitive
University Medical Centre,
study comparisons and is a barrier to research on risk impairment.12 However, neither of these guidelines
Amsterdam, Netherlands
(Prof F Barkhof MD); factors, pathophysiology, pathological correlations, and provide comprehensive recommendations for the
Department of Medicine, clinical consequences of these lesions. Indeed, the many forms of SVD seen on neuroimaging, and
Division of Neurology, Brain same lesions are classified differently across studies— neither include advances in understanding the
Research Centre, University of
eg, different definitions have resulted in small cavities pathophysiology and measurement of SVD, which are
British Columbia, Vancouver,
being classified as perivascular spaces or lacunes.9 changing rapidly.

822 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

Our international effort builds on previous initiatives

Variants of use of term Papers that use term
and aims to provide clear, rigorous, evidence-based, and in title or abstract*
easy-to-apply definitions and terminology for the (%) (n=1144)
structural neuroimaging features of SVD that avoid Leukoaraiosis Ischaemic leukoaraiosis, subcortical leukoaraiosis 350 (31%)
presumption of mechanisms of pathogenesis. We White matter lesions (WML) MRI white matter lesions, cerebral WML, T2 WML(s), 275 (24%)
include examples to help improve the standard use, cerebrovascular WML, subcortical WML, WML of
minimum advisory standards for image acquisition, Binswanger’s disease, cerebral WML of Binswanger’s
disease, confluent WML, intracranial WML
standards for analysis of imaging data on SVD and
White matter hyperintensity Cerebral WMH, age-related WMH, brain WMH, 217 (19%)
related features, and scientific reporting standards to (WMH) MRI WMH
improve clarity of publications on SVD.
White matter changes Age-related cerebral WMC, age-related WMC, 136 (12%)
Although this Position Paper focuses on the most (WMC) cerebral WMC, changes in white matter, age-related
common manifestations of SVD, other vascular lesions, changes in white matter
such as ischaemic or haemorrhagic stroke, subarachnoid Leukoencephalopathy Subcortical ischaemic leukoencephalopathy 76 (7%)
haemorrhage, subdural haematoma, and vascular mal- White matter disease (WMD) Age-related WMD, cerebral WMD, subcortical WMD 45 (4%)
formations unrelated to SVD, can also contribute to White matter damage Age-related white matter damage 5 (0%)
cognitive impairment and dementia, especially after Ischaemic white matter Ischaemic subcortical WMD, chronic ischaemic 4 (0%)
stroke. We provide a description of their features and disease cerebral WMD, subcortical ischaemic WMD
neuroimaging recommendations in the appendix. Others (9) ·· 17 (1%)
This Position Paper summarises the STandards for Data were derived from a structured literature search; for methodology and search strategy and selection criteria, see
ReportIng Vascular changes on nEuroimaging (STRIVE). appendix. *Instances that term was mentioned at least once in abstract or title. WML=white matter lesion.
Our main aim is to recommend standards for research WMH=white matter hyperintensity. WMC=white matter changes. WMD=white matter disease.
with MRI; however, many of the principles also apply to Table 1: Terms used to describe white matter hyperintensities of presumed vascular origin and
research with CT, and the standards might also facilitate frequency of use
a more consistent approach to identification of
manifestations of SVD on neuroimaging in clinical
practice. Small subcortical infarct Large subcortical infarct Other lesion (eg, small deep
(eg, striatocapsular) ICH, inflammatory lesion)
Methods
In 2011, the UK Medical Research Council (London, UK),
the German Centre for Neurodegenerative Disease
(DZNE, Bonn Germany), and the Canadian Institutes of
Health Research (Ottawa, ON, Canada) issued a call for Stroke symptoms,
proposals under a funding concordat of the Centres of no visible lesion
Excellence in Neurodegeneration (COEN), which aimed Acute
to accelerate progress in understanding the pathogenesis
of neurodegeneration.13 This initiative provided funding Time
for a working group of experts to establish standards for
neuroimaging in SVD. Chronic
In March, 2012, the core group of experts met in
Edinburgh, UK. Three study co-chairs (JMW, MDi, and
EES) identified experts from COEN-affiliated centres and
supplemented the group with representatives from
research groups active in neuroimaging of SVD from
non-COEN participant countries. Our group included
experts in neurology, neuroradiology, neuroepidemiology, Normal or nearly White matter Lacune
psychiatry, geriatrics, stroke, medical imaging physics, normal MRI hyperintensity

and neuropathology. The working method was based on Figure 1: Variable fates of lesions related to small vessel disease and the convergence of acute lesions with
the Delphi principle with workshops at the beginning different causes but similar late appearances on MRI
and end of the project, and with interim work Arrows indicate possible late fates of acute MRI findings. Blue arrows indicate common fates of recent small
subcortical infarcts, green arrows indicate less common fates, and red lines indicate least common late fates.
assignments (appendix). A template helped discussion to
ICH=intracranial haemorrhage.
focus on achievement of a consensus in key areas:
terminology, definitions, image acquisition, image
analysis, and reporting standards. We endorsed the affect future studies of the pathophysiology of SVD. BC, Canada
principle that terms and definitions should indicate Working groups developed standards for each of the six (Prof O R Benavente MD);
Sunnybrook Health Sciences
imaging characteristics as descriptively as possible, key lesion types, using established principles for Centre, Toronto, ON, Canada
avoiding presumptions of mechanism or pathological guideline development published by the Enhancing the (S E Black MD); Cambridge
links not well supported by published reports so as to not QUAlity and Transparency Of health Research (Equator) Institute of Public Health,

www.thelancet.com/neurology Vol 12 August 2013 823

 Position Paper

School of Clinical Medicine, Network. We did systematic searches to identify relevant Context, terminology, and definitions of
Cambridge, UK published work (appendix). imaging features
(Prof C Brayne PhD); German
Center for Neurodegenerative
The group reconvened in Munich, Germany, in Recent small subcortical infarct
diseases, Bonn, Germany November, 2012, to present draft standards from each Context
(Prof M Breteler MD); Service de working group for discussion and revision, with Clinically evident recent small subcortical infarcts,
Neurologie, Hopital additional review and comment from six new external commonly called lacunar strokes or lacunar syndrome,
Lariboisiere, INSERM,
Université Denis Diderot, Paris,
advisers. We wrote and revised the consensus document cause about 25% of all ischaemic strokes (figure 2).
France (Prof H Chabriat MD); with input from all workshop members. All participants Occasionally, a recent asymptomatic small subcortical
Department of Neurology, reviewed and endorsed the final document. infarct is identified by chance on imaging,14,15 and is
University of California at For the definition of SVD, we discussed size limits to referred to as a silent cerebral infarct. By contrast, for as
Davis, Sacramento, CA, USA
(Prof C DeCarli MD);
define perforating arteries and arterioles, but they were yet unknown reasons, in up to 30% of patients,
Department of Neurology, highly variable in published work, and did not translate symptomatic lacunar stroke syndromes seem not to be
Radboud University Nijmegen well to their appearance on imaging. Therefore, we accompanied by visible small subcortical infarcts,16
Medical Centre, Donders
decided to use the term arteriole to refer to small indicating that MRI is not fully sensitive in the detection
Institute for Brain, Cognition
and Behaviour, Nijmegen, perforating arteries and arterioles that are affected in of such infarcts. Additionally, some studies have shown
Netherlands (F-E de Leeuw MD); SVD. These standards are expected to reliably classify that the small subcortical infarcts might have differing
Institute for Stroke and most manifestations of SVD seen on neuroimaging; fates, evolving into a lacunar cavity or hyperintensity
Dementia Research, Klinikum
however, we acknowledge that individual judgment without apparent cavitation on T2-weighted sequences,
der Universität München,
Ludwig-Maximilians- might be needed for classification of ambiguous lesions or might disappear leaving little visible consequence on
Universität, Munich, Germany on the borders between categories (appendix), and that conventional MRI (figure 1). Estimates of the proportion
(M Duering MD, clinical judgment might be needed in some cases when of recent small subcortical infarcts that cavitate range
Prof M Dichgans MD);
using these standards in clinical practice. from 28%17 to 94%.18
Department of
Neurodegeneration, Dementia
Research Centre Recent small subcortical White matter Lacune Perivascular space Cerebral microbleed
(Prof N C Fox MD), Stroke infarct hyperintensity
Research Group, Department of
Brain Repair and Rehabilitation
(D Werring PhD), Institute of
Neurology, University College
London, London, UK;
Example image
Massachusetts General
Hospital, Stroke Research
Center, Boston, MA, USA
(Prof S Greenberg MD,
A Viswanathan MD);
Department of Clinical
Neurological Sciences, Western
University, London, ON,
Canada (Prof V Hachinski MD);
German Center for
T2
Neurodegenerative Diseases
Schematic
(DZNE) Rostock and
Greifswald, Rostock, Germany
(I Kilimann MD,
Prof S Teipel MD); Division of
Neurology, Department of DWI FLAIR FLAIR T1/FLAIR T2*/SWI
Medicine and Therapeutics,
Prince of Wales Hospital, Usual diameter ≤20 mm Variable 3–15 mm ≤2 mm ≤10 mm
Chinese University of Hong
Comment Best identified on DWI Located in white matter Usually have Most linear without Detected on GRE seq.,
Kong, Hong Kong Special hyperintense rim hyperintense rim round or ovoid, blooming
Administrative Region, China
(Prof V Mok MD); Department DWI ↑ ↔ ↔/(↓) ↔ ↔
of Neurology, School of Mental FLAIR ↑ ↑ ↓ ↓ ↔
Health and Neuroscience, and
Cardiovascular Research T2 ↑ ↑ ↑ ↑ ↔
Institute Maastricht, T1 ↓ ↔/(↓) ↓ ↓ ↔
Maastricht University Medical
Center, Maastricht, T2*-weighted GRE ↔ ↑ ↔ (↓ if haemorrhage) ↔ ↓↓
Netherlands
(Prof R van Oostenbrugge MD);
↑ Increased signal ↓ Decreased signal ↔ Iso-intense signal
Azienda Universitario
Ospedaliera Careggi,
Figure 2: MRI findings for lesions related to small vessel disease
Department of Neuroscience,
Shows examples (upper) and schematic representation (middle) of MRI features for changes related to small vessel disease, with a summary of imaging
Pharmacology and Child’s
characteristics (lower) for individual lesions. DWI=diffusion-weighted imaging. FLAIR=fluid-attenuated inversion recovery. SWI=susceptibility-weighted imaging.
Health (NEUROFARBA),
GRE=gradient-recalled echo.

824 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

Scarce data from pathological correlation studies University of Florence,

suggest that small subcortical infarcts are associated Panel 1: Glossary of proposed terms and definitions for Florence, Italy (L Pantoni MD);

with occlusion of small arteries, although the neuroimaging features of small vessel disease Department of Biomedical
Magnetic Resonance, Faculty
pathogenesis of these infarcts is unclear. Small Recent small subcortical infarct for Natural Sciences, Institute
subcortical infarcts occur in the perfusion territory of a Neuroimaging evidence of recent infarction in the territory of for Experimental Physics,
small artery or arteriole penetrating the internal part of Otto-von-Guericke
one perforating arteriole, with imaging features or clinical UniversityMagdeburg,
the brain. MRI studies suggest that recent small symptoms consistent with a lesion occurring in the previous Magdeburg, Germany
subcortical infarcts can exceed a diameter of 15 mm on few weeks. (Prof O Speck PhD); Newcastle
axial sections (the usual size limit for lacunes of University, Newcastle upon
presumed vascular origin) in the acute phase, and can Lacune of presumed vascular origin Tyne, UK (B C M Stephan PhD);
A round or ovoid, subcortical, fluid-filled cavity (signal similar Yong Loo Lin School of
be up to about 20 mm on axial sections. MRI also shows Medicine, National University
that small subcortical infarcts and lacunes can be more to CSF) of between 3 mm and about 15 mm in diameter,
of Singapore, Singapore,
than 20 mm long when measured in the coronal or consistent with a previous acute small subcortical infarct or Singapore (C Chen MD);
sagittal plane (appendix). haemorrhage in the territory of one perforating arteriole. Department of Radiology,
Leiden University Medical
White matter hyperintensity of presumed vascular origin Center, Leiden, Netherlands
Terminology Signal abnormality of variable size in the white matter that (Prof M van Buchem MD);
We sampled 641 abstracts and one previous systematic shows the following characteristics: hyperintensity on Department of Clinical
Sciences, Section of Neurology,
review9 and identified 159 different terms for recent T2-weighted images such as fluid-attenuated inversion Skåne University Hospital,
small subcortical infarcts; the most common terms were recovery, without cavitation (signal different from CSF). Lund, Sweden
lacunar infarcts, lacunar infarctions, and lacunar strokes Lesions in the subcortical grey matter or brainstem are not (Prof B Norrving MD); Saint
(appendix). We propose the new consensus term recent included in this category unless explicitly stated. If deep grey Mary’s Health Care, Hauenstein
Neuroscience Center, Grand
small subcortical infarct, removing the word lacunar matter and brainstem hyperintensities are also included, the Rapids, MI, USA
because of evidence that not all small subcortical infarcts collective term should be subcortical hyperintensities. (P B Gorelick MD); and Munich
become lacunes (ie, cavities).17,18 Cluster for Systems Neurology
Perivascular space (SyNergy), Munich, Germany
Definition Fluid-filled spaces that follow the typical course of a vessel as (Prof M Dichgans)

We propose that the term recent small subcortical it goes through grey or white matter. The spaces have signal Correspondence to:
intensity similar to CSF on all sequences (figure 2). Because Prof Joanna M Wardlaw, Division
infarct should refer to neuroimaging evidence of recent of Neuroimaging Sciences,
infarction in the territory of one perforating arteriole, they follow the course of penetrating vessels, they appear University of Edinburgh, Western
with imaging features or clinical symptoms consistent linear when imaged parallel to the course of the vessel, and General Hospital, Edinburgh,
with a lesion occurring in the previous few weeks round or ovoid, with a diameter generally smaller than 3 mm, EH4 2XU, UK
when imaged perpendicular to the course of the vessel. [email protected]
(figure 2, panel 1). Use of the word recent should refer
or
to lesions with symptoms or imaging features that Cerebral microbleed Prof M Dichgans, Institute for
suggest they occurred in the previous few weeks; the Small (generally 2–5 mm in diameter, but sometimes up to Stroke and Dementia Research,
word is used instead of acute because it includes the 10 mm) areas of signal void with associated blooming seen Klinikum der Universität,
first few weeks of the lesion, and not the hyperacute on T2*-weighted MRI or other sequences that are sensitive to Munich, Germany
stage only. Use of the word small indicates a lesion that [email protected]
susceptibility effects. muenchen.de
should be less than 20 mm in its maximum diameter in
the axial plane, although some lesions that appear to Brain atrophy See Online for appendix
represent infarction in the territory of one vessel can be A lower brain volume that is not related to a specific For more on Centres of
somewhat larger in the coronal plane (appendix).19 More macroscopic focal injury such as trauma or infarction. Thus, Excellence in
investigation is needed to precisely define upper size infarction is not included in this measure unless explicitly stated. Neurodegeneration see http://
www.coen.org
limits.
For more on the Equator
Lesions in the basal ganglia and internal capsule that Lacune of presumed vascular origin network please see http://www.
are larger than 20 mm and seem to be due to Context equator-network.org/
simultaneous infarction in several penetrating arteries Regarding pathological changes, Fisher21 wrote:
should not be classified as small subcortical infarcts, but “Historically, the original SVD feature was the lacune
rather as striatocapsular infarcts, a subtype of infarct (hole), which derived from French for a small fluid-filled
with a distinct cause.20 Similarly, infarcts of the anterior cavity that was thought to mark the healed stage of a small
choroidal artery are aetiologically distinct, identifiable by deep brain infarct. The term was adopted into English. By
their location (in the caudate nucleus head), and shape a process of medico-linguistic evolution, the precavitary
(mostly comma shaped), and therefore should not be phase became the lacunar infarct, the associated clinical
classified as small subcortical infarcts. Unlike lacunes of entity became the lacunar stroke and the neurological
presumed vascular origin, no lower size limit is given for features became the lacunar syndrome.” Lacunes are
small subcortical infarcts because diffusion-weighted frequently seen on imaging in elderly patients with no
imaging allows discrimination of small recent infarcts symptoms and are associated with an increased risk of
from perivascular spaces. stroke, gait impairment, and dementia.22–25 The cause of

www.thelancet.com/neurology Vol 12 August 2013 825

 Position Paper

most lacunes is presumed to be small subcortical infarcts, factors,29 the pathogenesis of these white matter lesions
either symptomatic or silent; however, some might result is not well understood and could be multifactorial.30
from small deep haemorrhages (figure 1).26 White matter hyperintensities are associated with covert
neurological and cognitive symptoms and physical
Terminology difficulties such as gait disturbance.31–35 Hyperintensities
Our systematic review identified more than 100 terms that can also occur in subcortical grey matter structures, such
have been used to describe lacunes of presumed vascular as the basal ganglia, and have sometimes been analysed
origin (appendix). Commonly used terms were lacune, alongside white matter hyperintensities. Hyperintensities
lacunar stroke, and silent brain infarct. We propose the can also be present in the brainstem. Some investigators
new term lacune of presumed vascular origin, which have differentiated between hyperintensities of
discriminates between small cavitated lesions of presumed periventricular and deep white matter (appendix), with the
vascular origin and other small brain cavities, and allows suggestion, under debate, that they have differing
for some uncertainty about the ischaemic or haemorrhagic pathogenesis, risk factors, and clinical consequences.
origin of the lesion when no imaging is available in the Many investigators have included total white matter
acute phase, as is commonly the case (panel 1). hyperintensities in their analyses.

Definitions Terminology
We define a lacune of presumed vascular origin as a Our systematic review of 940 abstracts identified
round or ovoid, subcortical, fluid-filled (similar signal as 50 different terms for white matter hyperintensity. The
CSF) cavity, of between 3 mm and about 15 mm in most common terms were leukoaraiosis, white matter
diameter, consistent with a previous acute small deep lesions, white matter hyperintensities, leukoencephal-
brain infarct or haemorrhage in the territory of one opathy (usually in the context of cerebral autosomal
perforating arteriole (panel 1). On fluid-attenuated dominant arteriopathy with subcortical infarcts and
inversion recovery (FLAIR) images, lacunes of presumed leukoencephalopathy [CADASIL]), and white matter
vascular origin generally have a central CSF-like disease (table 1; appendix).
hypointensity with a surrounding rim of hyperintensity; Leukoaraiosis describes a reduced area of x-ray
however, the rim is not always present, and a hyperintense attenuation on CT,36 and this term was later adopted to
rim can also surround perivascular spaces when they denote hyperintensity on T2-weighted and FLAIR MRI,
pass through an area of white matter hyperintensity and sometimes also hypointensity on T1-weighted MRI
(figure 2). In some cases, the central cavity fluid is not (appendix), a situation in which the radiological
suppressed on FLAIR, and the lesion can appear entirely description came before the pathological description.
hyperintense, despite MRI having a clear CSF-like We propose the term white matter hyperintensity of
intensity on other sequences such as T1-weighted and presumed vascular origin to exclude white matter lesions
T2-weighted MRI.18 from other diseases such as multiple sclerosis or
Lacunes of presumed vascular origin should be leukodystrophies.
distinguished from perivascular spaces. Although
pathological studies have not shown an absolute cutoff Definitions
size, lesions that are less than 3 mm in diameter are White matter hyperintensities of presumed vascular
more likely to be perivascular spaces than to be lacunes origin are hyperintense on T2-weighted sequences and
(appendix), and we recommend the use of this size can appear as isointense or hypointense (although not as
criterion to discriminate between the two lesions, which hypointense as CSF) on T1-weighted sequences,
is consistent with previous studies.27,28 We chose a depending on the sequence parameters and severity of
maximum size of 15 mm for lacunes of presumed pathological change (figure 2, panel 1; appendix).
vascular origin—which differs from the 20 mm diameter Members of our group had differing opinions about
of recent small subcortical infarcts—since old infarcts whether grey matter hyperintensities or brainstem
are generally smaller than recent infarcts because of hyperintensities should be routinely classified as white
tissue loss and an ex-vacuo effect in old lesions, and matter hyperintensities, as some previous studies have
because of swelling in new lesions. However, we done. The consensus opinion was that lesions in the
recognise that this size boundary is not supported by subcortical grey matter or brainstem should not be
much objective evidence, and research is needed. included in the category of white matter hyperintensity of
presumed vascular origin unless explicitly stated; we
White matter hyperintensity of presumed vascular origin endorsed subcortical hyperintensities as an acceptable
Context alternative collective term for any non-cortical
White matter lesions characterised by bilateral, mostly hyperintensities, including those in white matter, deep
symmetrical hyperintensities on T2-weighted MRI are grey matter, and the brainstem. When using CT, white
common in older individuals. Although strongly matter hypoattenuation or white matter hypodensities can
associated with cerebrovascular disease and vascular risk be used because of the appearance of the lesions on CT.

826 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

Perivascular space from small lacunes of presumed vascular origin

Context (appendix). By contrast with lacunes, the diameter of
Perivascular spaces are extensions of the extracerebral perivascular spaces is not usually more than 3 mm as
fluid space around arteries, arterioles, veins, and venules confirmed pathologically,49,50 and spaces do not have a
as they course from the brain surface into and through T2-hyperintense rim around the fluid-filled space on T2-
the brain parenchyma, and they can be followed by sheets weighted or FLAIR imaging, unless they traverse an
of leptomeninges.37 Perivascular spaces are commonly area of white matter hyperintensity.51
microscopic, and not visible on conventional neuro-
imaging; however, larger spaces become increasingly Cerebral microbleed
apparent with increasing patient age, especially when Context
located at the base of the brain.38 General enlargement of Cerebral microbleeds are small hypointense lesions that
perivascular spaces is associated with other morphological are visible on paramagnetic-sensitive MR sequences
features of SVD such as white matter hyperintensities39 such as T2*-weighted gradient-recalled echo (GRE) or
and lacunes,40 but not atrophy.41 Whether the presence of susceptibility-weighted sequences, and are most
several visible perivascular spaces is clinically significant commonly located in the cortico-subcortical junction,
remains controversial, and the spaces should therefore and deep grey or white matter in the cerebral
not be referred to as lesions; however, some studies have hemispheres, brainstem, and cerebellum (figure 2).
associated more prominent perivascular spaces with Results of some studies30,52,53 suggest that MR-visible
worse cognitive function.42 lesions correspond to haemosiderin-laden macrophages
in perivascular tissue, consistent with vascular leakage of
Terminology blood cells.52,54–56 Cerebral microbleeds are associated with
Synonymous terms for perivascular spaces include SVD and Alzheimer’s disease.57,58 The presence or
Virchow–Robin spaces,40 type 3 lacune,43 or état crible when absence of strictly lobar microbleeds has been included
located predominantly in the basal ganglia.38 Terms used to in research criteria for cerebral amyloid angiopathy.59
describe visible perivascular spaces include large, dilated, Originally thought to be asymptomatic markers of SVD,
large dilated, or enlarged Virchow–Robin spaces or emerging data shows association between microbleeds
perivascular spaces.38,41,42,44–48 We did not consider these and cognitive impairment, although the mechanisms of
terms appropriate because the association between the size this association, including whether microbleeds damage
of the perivascular spaces and clinical consequences is not the brain and cause dysfunction, are not well understood.
well understood, and because the visibility of the spaces Recommendations for imaging and reporting of
depends on MRI sequence characteristics, which vary microbleeds have been published,60 and we endorse the
across studies. Visibility alone cannot be a uniform criterion use of these standards. Amyloid-related imaging
for pathologically enlarged perivascular spaces. We abnormalities–haemorrhage (ARIA-H) describes
therefore recommend the consensus term perivascular microbleeds in the context of Alzheimer’s disease or
space. β-amyloid immunotherapies. A consensus group from
the US Alzheimer’s Association has published standards
Definitions for assessing microbleeds in this specific clinical
We define perivascular spaces as fluid-filled spaces that context,61 and we also endorse the use of these guidelines.
follow the typical course of a vessel as it goes through
grey or white matter. The spaces have signal intensity Terminology
similar to that of CSF on all sequences (figure 2); Our systematic review revealed 387 instances of 20 different
because they follow the course of penetrating vessels, terms for microbleeds in 370 abstracts (appendix).
they appear linear when imaged parallel to the course of Microbleed was by far the most common term, followed by
the vessel, and round or ovoid, with a diameter generally cerebral microbleed, then brain microbleed. We propose
smaller than 3 mm, when imaged perpendicular to the the consensus term cerebral microbleed.
course of the vessel (panel 1, appendix). At high
resolution, a central vessel can occasionally be seen in Definitions
the centre of a perivascular space, which could possibly Cerebral microbleeds are visualised as small
differentiate the spaces from lacunes. Perivascular (generally 2–5 mm in diameter, but up to 10 mm) areas of
spaces are generally most prominent in the inferior signal void with associated blooming seen on T2*-weighted
basal ganglia, and can also be seen coursing centripetally MRI or other sequences that are sensitive to susceptibility
through the hemispheric white matter and in the effects (panel 1),57,60 and are generally not seen on CT, or on
midbrain; however, the spaces are rarely seen in the FLAIR, T1-weighted, or T2-weighted sequences. When
cerebellum. Perivascular spaces can show focal imaged with T2*-weighted GRE sequences, cerebral
enlargement, and can be especially enlarged (up microbleeds are well defined, of homogeneous low signal,
to 10–20 mm, even with mass effect) in the inferior and are either round or oval in shape (figure 2). Because
basal ganglia. Perivascular spaces must be discriminated the blooming artifact, and therefore the visualised size,

www.thelancet.com/neurology Vol 12 August 2013 827

 Position Paper

depends on MR field strength and sequence, we do not haemorrhages, which might be caused by cerebral
recommend an absolute criterion for size. When imaged amyloid angiopathy, should be distinguished from non-
with 1·5 T and 3·0 T GRE sequences, cerebral microbleeds lobar intracerebral haemorrhages, which are not thought
are generally 2–5 mm in diameter, but can be up to 10 mm. to be caused by cerebral amyloid angiopathy, and are
Hypointensities of less than 2 mm, which could be mostly due to perforating arteriolar vasculopathy.64,65
attributable to signal loss from only one voxel, should be Results of a small study66 suggested good, but not perfect,
regarded as questionable on 1·5 T MRI because such small inter-rater agreement on the site of origin of intracerebral
hypointensities could be artifacts. haemorrhages. The Boston criteria59 for cerebral amyloid
Susceptibility-weighted imaging can also be used to angiopathy should be used in assigning the likelihood of
assess cerebral microbleeds. New quantitative-based underlying cerebral amyloid angiopathy in patients with
methods (ie, quantitative susceptibility mapping) need lobar intracerebral haemorrhages.
more investigation, but might improve the assessment We propose the term superficial cortical siderosis for
of cerebral microbleeds. Several lesions or structures neuroimaging evidence of chronic blood products in the
can mimic cerebral microbleeds, such as calcification, superficial cortex under the pia mater. Superficial
normal vessels seen in cross-section, iron deposits cortical siderosis could be a chronic consequence of
from other causes, haemorrhagic metastases (eg, subarachnoid bleeding or might be the result of very
melanoma), and diffuse axonal injury (eg, after head superficial cortical bleeding caused by vascular
trauma).60 Cerebral microbleeds can be differentiated malformations, cerebral amyloid angiopathy,67 spinal
from an old small deep spontaneous intracerebral dural defects, or it might also be idiopathic.67,68
haemorrhage because, in general, the intracerebral T2*-weighted GRE or other blood-sensitive sequences
haemorrhages are larger, irregular with a cystic cavity will show superficial cortical siderosis as a linear
(figure 1), and will be visible on T1-weighted and hypointensity over the cortex, but this disorder might be
T2-weighted or FLAIR sequences. mimicked by the petechial cortical haemorrhagic
transformation of an infarct. Revised research criteria
Other haemorrhagic lesions for cerebral amyloid angiopathy include superficial
We discussed two other haemorrhagic manifestations of cortical siderosis as an additional haemorrhagic
SVD: intracerebral haemorrhage and superficial cortical manifestation of cerebral amyloid angiopathy,
siderosis (appendix). Intracerebral haemorrhage could equivalent to a lobar intracerebral haemorrhage.67
be a manifestation of SVD62 or secondary to other causes Investigators should describe the location of siderosis
such as vascular malformations. We suggest the use of (ie, the number of sulci involved67 and in which lobes).
the consensus term spontaneous intracerebral
haemorrhage presumed to be due to SVD, instead of Brain atrophy
secondary intracerebral haemorrhage from other causes Context
or traumatic intracerebral haemorrhage. Comprehensive Brain atrophy can be general or focal (affecting only
guidelines for the diagnosis and management of particular lobes or specific brain regions—eg, the
intracerebral haemorrhage have been published hippocampus), symmetrical or asymmetrical, or tissue
previously.63 In the context of SVD, we encourage selective (affecting a certain tissue class—eg, white
researchers to resolve the contribution of different matter), and occurs in many disorders. The pathological
vascular diseases to intracerebral haemorrhage. In the changes of atrophy are heterogeneous and not
meantime, we recommend that lobar intracerebral necessarily indicative of neuronal loss.69–71 Brain atrophy
occurs with the usual ageing process, but the extent
Baseline 36 months varies between individuals. In the context of vascular
disease and dementia, neuropathological substrates of
atrophy include neuronal loss,72 cortical thinning,
subcortical vascular pathology with white matter
A
rarefaction and shrinkage, arteriolosclerosis, venous
collagenosis, and secondary neurodegenerative
changes.73,74 Many imaging studies report an association
between the presence and severity of SVD and brain
atrophy, including global atrophy, corpus callosum
atrophy, central atrophy (increased ventricular size and
atrophy of the basal ganglia), mesencephalic atrophy,
B
D C and hippocampal atrophy, and focal cortical thinning in
brain regions connected to subcortical infarcts
Figure 3: Secondary brain atrophy in a 55-year-old patient with documented small vessel disease
Baseline (middle). The follow-up scan (T1-weighted MRI; right) shows clear sulcal widening (arrow B, C, and D),
(figure 3).75,76 Therefore, vascular lesions should be
particularly in occipital regions, and ventricular enlargement (arrow A) without new infarctions during the included in studies of atrophy; conversely, atrophy is an
observational period. Fluid-attenuated inversion recovery image (left) shows substantial white matter hyperintensity. important measure in imaging studies that are done to

828 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

Purpose* Orientation Target-slice Comment

thickness and in-
plane resolution
Minimum essential sequences—eg, for clinical or large-scale epidemiological studies, available on most MRI scanners
T1-weighted Important for discriminating lacunes from dilated perivascular 2D axial, sagittal, 3–5 mm, and 1 mm At least one sequence in sagittal or coronal plane is
spaces; for discriminating grey from white matter, and for or coronal × 1 mm helpful to visualise full extent and orientation of lesions
studying brain atrophy
DWI The most sensitive sequences for acute ischaemic lesions; 2D axial 3–5 mm, and 2 mm Reduced signal on apparent diffusion coefficient map
positive for up to several weeks after cerebrovascular event × 2 mm helps to discriminate recent lesions from old lesions
T2-weighted To characterise brain structure; to differentiate lacunes from 2D axial 3–5 mm, and 1 mm ··
white matter hyperintensities and perivascular spaces; to × 1 mm
identify old infarcts
FLAIR To identify white matter hyperintensities and established 2D axial 3–5 mm, and 1 mm ··
cortical or large subcortical infarcts; to differentiate white × 1 mm
matter lesions from perivascular spaces and lacunes
T2*-weighted GRE To detect haemorrhage, cerebral microbleeds, siderosis; for 2D axial 3–5 mm, and 1 mm Only reliable routine sequence for detection of
measurement of intracranial volume × 1 mm haemorrhage
Other routine sequences, available on most MR scanners
Proton density-weighted To detect white matter hyperintensities, infarcts, perivascular 2D axial 3–5 mm, and 2 mm Mostly replaced by FLAIR
spaces (with T2-weighted dual echo), or other pathologies × 2 mm
MRA To detect stenosis of vertebral, basilar, internal carotid, middle Post-contrast or 3D, axial, coronal, Only large vessels visible at 1·5 T or 3·0 T; see below for
cerebral, anterior cerebral, or posterior cerebral artery, or other 3D time-of-flight sagittal perforating arterioles
pathologies for intracranial reconstruction;
arteries 1 mm isotropic
voxels
Sequences commonly available on commercial clinical MR scanners; at present, used more for research studies, but some techniques are increasingly used in clinical protocols
DTI with six-gradient To diagnose recent infarct; measurement of mean diffusivity 2D axial 3–5 mm, and 2 mm More detailed characterisation than with DWI;
direction diffusion and fractional anisotropy × 2 mm acquisition time is double that for DWI
encoding
SWI or equivalent Very sensitive to haemosiderin, measurement of intracranial 2D or 3D axial 2D: 3–5 mm, and Enables visualisation of more cerebral microbleeds than
volume 2 mm × 2 mm; T2*-weighted GRE imaging and is more sensitive to
3D: 1 mm isotropic artifacts including motion
voxels
Research-only sequences; require research expertise
Isotropic volumetric T2- To display fine detail of perivascular spaces 3D axial 1 mm isotropic Allows post-acquisition reformatting; could potentially
weighted voxels replace 2D T2-weighted imaging if signal-to-noise ratio
is adequate
Isotropic volumetric 3D T1- Provides improved global and regional volumetric brain 3D axial 1 mm isotropic Allows post-acquisition reformatting; could potentially
weighted (eg, MP-RAGE) measurements voxels replace 2D T1-weighted imaging if signal-to-noise ratio
is adequate
Isotropic volumetric FLAIR Enables identification of white matter hyperintensities; used 3D axial 1 mm isotropic Allows post-acquisition reformatting; could potentially
for imaging cortical or subcortical infarcts voxels replace 2D FLAIR imaging if signal-to-noise ratio is
adequate; more homogeneous CSF suppression
Advanced DTI with more Provides refined and superior quantitative measurements of 2D axial 3–5 mm, and Allows for tractography, connectome mapping, and
than six-direction diffusion microscopic tissue changes 2 mm × 2 mm more accurate measurements of mean diffusivity and
encoding (eg, 32 or more fractional anisotropy
diffusion-encoding
directions)
MTR To detect demyelination and axonal loss 2D axial 3–5 mm, and 1 mm Experience in acquisition and interpretation needed;
× 1 mm involves two measurements (with and without
magnetisation transfer-pulse)
T1 mapping To measure water content of tissue Axial 3–5 mm, and 2 mm Experience in acquisition and interpretation needed
× 2 mm
Permeability imaging To estimate permeability of the blood–brain barrier Axial; sequential 3–5 mm, and 2 mm Intravenous contrast injection needed; involves
before and after × 2 mm complex image processing; methods improving rapidly
contrast
ASL perfusion imaging To measure tissue perfusion; quantitative, with assumptions 2D axial 3–5 mm, and 2 mm Complex to set up and run accurately; needs post-
× 2 mm processing; optimum processing strategies not yet
confirmed; contrast injection not needed
Perfusion imaging (DCE or To semiquantitatively measure blood perfusion in tissue 2D axial 3–5 mm, and 2 mm Needs intravenous injection of contrast agent and
DSC) × 2 mm post-processing; optimum acquisition and processing
not yet confirmed for T1 (DCE) or T2*-weighted (DSC)
approaches
(Continues on next page)

www.thelancet.com/neurology Vol 12 August 2013 829

 Position Paper

Purpose* Orientation Target-slice Comment

thickness and in-
plane resolution
(Continued from previous page)
fMRI To measure brain function in response to tasks or stimuli, or at 2D axial 3–5 mm, and 2 mm Complex set-up, acquisition, and processing
rest for default mode networks × 2 mm
QSM To provide quantitative measures of susceptibility changes, 2D or 3D axial 2D: 3–5 mm, and Uses an SWI-like acquisition, but needs very complex
independent of scanner or acquisition variables 2 mm × 2 mm; post-processing methods; post-processing strategies
3D: 1 mm isotropic currently under investigation
voxels
Microatheroma and To visualise perforating arteriolar anatomy and atheroma Uncertain, Uncertain, Promising experimental approach that needs a scanner
arteriolar imaging emerging method emerging method that is more than 3·0 T

DWI=diffusion-weighted imaging. FLAIR=fluid-attenuated inversion recovery. GRE= gradient-recalled echo. MRA=magnetic resonance angiography. DTI=diffusion tensor imaging. SWI=susceptibility-weighted
imaging. MP-RAGE=magnetisation-prepared rapid acquisition with gradient echo. MTR=magnetisation transfer ratio. ASL=arterial spin labelling. DCE=dynamic contrast-enhancement. DSC=dynamic
susceptibility contrast. fMRI=functional MRI. QSM=quantitative susceptibility mapping. *MRI at 3·0 T is preferred to 1·5 T. However, these standards are listed as minimum and essential to research-only
applications. These categories are not absolute; purposes are variable, and will vary with investigators’ interest, expertise, and available technology.

Table 2: Proposed image acquisition standards for neuroimaging of small vessel disease

assess the burden of vascular damage in the brain, and SVD, intracerebral haemorrhage from other causes,
atrophy is thought to mediate, at least partially, the subarachnoid haemorrhage, subdural haematoma,
effects of vascular lesions on cognition.77–79 vascular malformations, and large artery ischaemic dis-
ease. These features are discussed in detail in the appendix.
Terminology
Our systematic review of studies on SVD and atrophy Advised minimum standards and parameters
revealed the use of many synonymous terms, including for imaging of SVD
atrophy, brain volume, volume loss, and others. We Image acquisition
propose the consensus term brain atrophy. We also If no contraindications are known, MRI, rather than CT,
propose that studies should use unambiguous terms is preferred for research and routine clinical use because
about whether the atrophy assessment was cross- it has higher sensitivity and specificity for detecting most
sectional or longitudinal, and should avoid terms such as manifestations of SVD (table 2). A field strength of 3·0 T
accelerated atrophy or brain volume loss in cross- might be preferred, but images from modern 1·5 T MRI
sectional studies in favour of more atrophy or lower brain scanners are often of a similarly high definition,80 and are
volume, respectively. When discussing atrophy, therefore acceptable; 1·5 T systems are also more widely
investigators should be clear about the specific brain available than 3·0 T systems, an important consideration
volumes they have measured and should include the for multicentre studies. Imaging sequences should be
brain subregion in their description—eg, hippocampal optimised for field strength and scanner configuration. If
atrophy. CT is used to image SVD, investigators should obtain a
thin-section volume sequence, without contrast, covering
Definitions the whole brain, and using a brain parenchymal
We define brain atrophy in the context of SVD on imaging algorithm and axial 5 mm reconstructions. However, CT
as a lower brain volume that is not related to a specific is no longer recommended, except possibly in large-scale
macroscopic focal injury such as trauma or infarction epidemiological studies, or if MRI is not available or is
(panel 1). Tissue loss is assumed from the enlargement of too expensive.
peripheral (sulcal) and central (ventricular) CSF spaces in For MRI, the minimum acceptable examination
relation to intracranial volume and other measures should include axial diffusion-weighted imaging (and
(figure 3). However, ideally, such assumptions should be apparent diffusion coefficient map), FLAIR,
formally confirmed with longitudinal observation. Tissue T2-weighted, and T2*-weighted GRE or susceptibility-
loss from discrete focal lesions, such as cortical infarcts, is weighted imaging, and T1-weighted imaging. The
easier to see on cross-sectional imaging, and should not be diffusion-weighted imaging sequence is very important
confused with generalised global or regional brain atrophy for identification of recent infarcts—eg, when used in
that is probably secondary to a diffuse process. patients with symptoms. MRI with diffusion-weighted
imaging should be considered the reference standard
Other vascular lesions for recent small subcortical infarcts, although results
These consensus standards focus on SVD. However, other might be falsely negative in some cases.16 The
vascular lesions can also be seen on imaging, particularly combination of clinical syndrome plus MRI results but
in elderly individuals and in those with cognitive without diffusion-weighted imaging, or clinical
impairment. These include ischaemic lesions unrelated to syndrome plus CT results, is not optimal; clinical

830 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

syndrome alone is the least reliable method of sequence also allows anatomical coregistration and
assessment.81 The use of one sequence (usually quantification of brain volume. The Alzheimer’s
T1-weighted MRI) in the coronal or sagittal plane Disease Neuroimaging Initiative has validated three-
instead of the axial plane helps to visualise the correct dimensional magnetisation prepared T1-weighted
lesion dimensions (appendix). Slice thickness should sequence protocols for the major scanner types at 1·5 T
be 5 mm or less, ideally with no gap; in-plane resolution and 3·0 T; sequence protocols are freely available. For more on the Alzheimer’s
should be 1 mm by 1 mm or better. Whole-brain For a comprehensive research MRI protocol for SVD, Disease Neuroimaging
Initiative please see http://www.
coverage is recommended for all sequences to achieve investigators should consider use of higher-resolution adni-info.org
optimum image analysis. Three-dimensional T1- MRI than the sequences specified above, and use volume
weighted thin-section isotropic sequences are now acquisitions to allow for volumetric whole-brain imaging
widely available and quick and can be reformatted on with near-isotropic resolution. These three-dimensional
thicker slices for ease of viewing in multiple planes; this acquisitions are now possible on most scanners for

Measures of interest Qualitative analysis standards Quantitative analysis Study design Accuracy, reliability, General comment
standards feasibility
Recent small Number (multiplicity Various coding schemes available for Possible, but Cross-sectional and Easy to identify on DWI, Mimics include acute
subcortical might indicate other location: anatomical (eg, centrum impractical for size and longitudinal: recent small reliability depends on time inflammatory multiple
infarct causes);85,86 size semiovale, corona radiata, basal volume subcortical infarcts are between infarct and sclerosis plaques; acute
(maximum diameter); ganglia, thalamus, internal capsule, typically detected in the imaging; more difficult lesions generally have
volume; location external capsule, optic radiation, setting of an acute clinical when using other increased signal on DWI
(anatomical region, cerebellum, and brainstem), and the event, but can also be an sequences or CT without and reduced signal on
vascular territory); shape vascular territory (eg, middle cerebral incidental finding longitudinal data apparent diffusion
(round, ovoid, tubular); artery, posterior cerebral artery, coefficient images
swelling (indicates recent, internal carotid artery, and basilar
not old) artery)
Lacune of Number (one or many); Various coding schemes available for Protocols for Cross-sectional and Differentiation from Hypointense rim on
presumed size (maximum shape and location: anatomical quantitative longitudinal: particular perivascular spaces can be T2*-weighted imaging
vascular origin diameter); shape (round, (eg, lentiform nucleus, thalamus, measurement care is needed to difficult; high observer suggests previous small
ovoid, tubular, other); internal capsule, centrum semiovale, available, need manual differentiate lacunes from agreement should be deep haemorrhage
location (anatomical brainstem); prominent ex-vacuo correction perivascular spaces; achieved before
region); evidence of effect indicates lesion was originally longitudinal: difference in undertaking actual ratings
previous haemorrhage; larger (eg, striatocapsular infarct)20 imaging helps to identify
ex-vacuo effect incident lacunes
White matter Volume; location Various coding schemes available for Various visual rating Cross-sectional and Inter-rater and intra-rater Careful visual checking is
hyperintensity (anatomical region); anatomical location scores87–92 and protocols longitudinal: consider reliability for both needed at all stages of
number (eg, periventricular, for quantitative masking recent small qualitative and computational analysis to
deep, subcortical, brainstem; or measurement93,94–97 are subcortical lesions, quantitative analysis of avoid difficulties from
centrum semiovale, corona radiata, available; the two lacunes, and perivascular white matter excess lesion distortion by,
internal capsule, external approaches are spaces when measuring hyperintensity is high if for example, bias field
capsule, optic radiation, brainstem; or complementary;82,83 volume of white matter done by trained raters, correction; regular
frontal, temporal, parietal, occipital) outputs should be hyperintensity to avoid with intraclass correlation recalibration against
visually reviewed by an inflating the volume; coefficients generally standard examples is
experienced rater for longitudinal: difference above 0·90; visual rating needed in rating large
mimics, artifacts, focal imaging might help to scores might have ceiling numbers of scans
infarcts, and identify new white matter or floor effect so
mislabelling lesions performance can differ
with extent of disease
Perivascular Number (multiplicity); Anatomical: midbrain, hippocampus, Visual scores used to Cross-sectional: consider Difficult to determine, Can be difficult to
space location (anatomical basal ganglia, centrum semiovale rate number of lesions masking perivascular especially when numerous, distinguish from lacunes;
region); size (maximum in basal ganglia, spaces when measuring and in the presence of giant perivascular spaces
diameter); shape centrum semiovale, volume of white matter white matter can be greater than 2 cm,
midbrain;42,45,46 various hyperintensity, although hyperintensities and are most commonly
threshold-based this might be difficult; located below the
methods are in longitudinal: little putamen
development experience
Cerebral Number (few or Some semi-automated approaches Several visual scores are Cross-sectional and Inter-rater agreement for Lobar and deep cerebral
microbleed multiple); location (lobar, segment cerebral microbleeds as an available;60,100,101 no longitudinal: consider use the presence or absence of microbleeds might have
deep, or infratentorial; extra tissue class or radial symmetry methods available for of visual scores; one or two microbleeds different risk factors and
anatomical region); size and mask areas of mineralisation,98,99 automated detection longitudinal: no specific varies, but agreement causes (eg, lobar cerebral
but these are experimental at scores available for (ie, 0·8) between the microbleeds are associated
present, and need validation longitudinal studies numbers of microbleeds is with cerebral amyloid
reasonable; reliability can angiopathy)
be improved through the
use of standardised
scales60,100.101
(Continues on next page)

www.thelancet.com/neurology Vol 12 August 2013 831

 Position Paper

Measures of interest Qualitative analysis standards Quantitative analysis Study design Accuracy, reliability, General comment
standards feasibility
(Continued from previous page)
Brain atrophy Whole brain should be If scans are not suitable for Automated or semi- Cross-sectional: brain Computational approaches Consider masking recent
adjusted for intracranial volumetric techniques or if such automated atrophy can be estimated have high reliability; visual small subcortical lesions,
volume; regional techniques are not available, quantitative methods by comparison with the rating is more varied but lacunes, and perivascular
(hippocampus, specific qualitative rating scales could provide are preferred but visual inner-skull volume (an can be improved with space when measuring
gyri, lobes should be an alternative102,103 checking and manual estimate of maximum reference to a standard brain volume;78,105 specific
adjusted for whole-brain editing are commonly brain size in patients at visual template;103 standards are emerging
volume); cortical or needed to avoid around age 20 years); all subcortical and cortical for hippocampal volume
subcortical; superficial or including the orbits and intracranial contents must vascular lesions affect the measurement107
deep (sulcal or ventricular excluding the be included in the reliability of automated
enlargement; whole- brainstem from the intracranial volume, volumetric techniques,79
brain volume adjustment whole-brain including veins and particularly in subjects with
needed) volume;69,93,104 regional meninges, which expand a high lesion load
or subregional brain into space left by shrinking
volume computational brain;76 longitudinal: serial
methods are in brain volumes can be
development, but their measured; a registration-
reliability, especially in based approach is
individuals with preferred, although the
disease, is still to be discipline is advancing
determined78,105 rapidly106

DWI=diffusion-weighted imaging.

Table 3: Proposed analysis standards for neuroimaging features of small vessel disease

T1-weighted, T2-weighted, GRE, and FLAIR techniques. with more than one rater, inter-rater reliability should be
Other high-resolution sequences sensitive to paramagnetic assessed. Observers should have good knowledge of
content, such as susceptibility-weighted imaging and normal neuroanatomy to measure global and regional
equivalents, should be considered for three-dimensional atrophy. Observers should train on established training
image acquisition, although these sequences increase the sets, and should aim for high observer agreement before
presence of artifacts. Diffusion tensor imaging builds on taking actual measurements. Regular recalibration against
findings with diffusion-weighted imaging, and should be standard examples is useful to maintain consistency when
used in research protocols. Diffusion tensor imaging is rating large numbers of scans or when rating images from
useful in the assessment of surrogates for different sessions.
microarchitectural integrity in normal-appearing and When investigators use automated computational
lesional white matter, in the determination of structural methods for analysis, we emphasise that the accuracy of
connectivity of white matter, and in tractography. For each individual’s results should be visually confirmed by
quantitative volumetric assessments of atrophy, we advise a trained observer. This validation step is crucial for
a high-resolution (1·0 mm to 1·5 mm isotropic voxels) T1- studies of SVD because automated methods have not
weighted sequence with good grey–white matter been developed for the disorder, and there is a high risk
differentiation. Other MR measurements show promise of misclassification of lesions, especially when unfamiliar
for imaging of SVD in research, including measurements pulse sequences are used or scanners are upgraded.
of perfusion, magnetisation transfer, blood–brain barrier Known examples to avoid are the misclassification of
permeability, vascular reactivity, metabolites, and white matter hyperintensities as normal grey matter, of
microatheroma in perforating arterioles. lacunes as ventricle or white matter hyperintensity, or of
small subcortical infarcts as white matter
Image analysis hyperintensities.84 For several of these misclassifications,
Qualitative (visual rating) and quantitative (computational manual correction will be needed.78,85
analysis) methods are established or are emerging for all For small numbers of discrete lesions, such as recent
lesion types. For some of the lesions, qualitative and small subcortical infarcts, lacunes, and microbleeds, we
quantitative approaches are closely related,82,83 and both recommend that the location of each lesion and the total
have advantages and disadvantages that should be number of lesions are recorded. At minimum, we
considered in the design of any study of SVD (appendix). recommend recording of the number of lesions in each
For all lesion types, observers should be trained in of the following brain subregions, on both the right and
neuroimaging interpretation, and specifically how to left sides: cerebral cortex (divided by lobe), corona radiata
recognise lesions of SVD. If possible, particularly in trials and centrum semiovale, putamen, globus pallidus,
or large observational studies, images should be rated thalamus, internal capsule, external capsule, brainstem,
centrally by one or a few experienced raters. In studies and cerebellum (table 3, appendix).

832 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

Reporting standards for vascular findings on rating scales should use established and widespread
neuroimaging rating scales so that studies are comparable and meta-
Vascular findings from neuroimaging should be reported analyses can be done.
according to standards established by the Equator
Network for observational studies, epidemiological Future developments and challenges
studies, diagnostic test assessments, or randomised Advances in the technology for imaging, as well as in novel
clinical trials. Additionally, we suggest specific points protocols for image acquisition and post-processing, have
that will help in the progression and standardisation of improved imaging of the various manifestations of SVD.
future research with imaging (panel 2). We encourage These advances have the potential to help explain the role
investigators to use the terms and definitions proposed of SVD in neurodegenerative disease and to identify new
in future studies. Where possible, studies with visual mechanisms of disease. Examples of promising technology

Panel 2: Proposed reporting standards for neuroimaging studies of small vessel disease

Demographic details of the research participants and (eg, neurology, psychiatry, neuroradiology, or radiology)
reference population and experience; rater reliability (intra-rater and inter-rater).
• Proportions of individuals with vascular risk factors and how • For studies using computational image analysis
these were measured. programmes: training of the analysts, any expert
• Proportions of individuals with stroke and its subtypes, as supervision, and the background of the expert; repeatability.
well as other vascular disease. • Statistical methods used in data analysis.
• Time from disease presentation to imaging and clinical • Ideally: sample size estimation.
assessments (if relevant).
Small vessel disease-specific aspects
• Any clinical or imaging observational period with time
• For recent small subcortical infarcts: specify whether infarcts
intervals.
are symptomatic or not; state the location, size, shape, and
• For studies on cognition or specific physical functions:
number; specify the delay from stroke to imaging; state the
details of test versions used, who administered them, and
proportion with visible acute lesion on diffusion-weighted
their training.
imaging and fluid-attenuated inversion recovery, plus
• For cognitive studies: assessment of premorbid cognitive
T2-weighted imaging.
ability and depression.
• For lacunes of presumed vascular origin: specify location,
Image acquisition size, shape, and number; distinguish haemorrhagic lesions
• Scanner characteristics (type and manufacturer, field from lacunes; for volumetric methods, state whether lacunes
strength, coils, high-order shim and use of shimming are counted as part of CSF volume, as part of the white
routines, quality assurance protocol for scanner and matter hyperintensity volume, or as separate lacune volume.
frequency of quality assurance assessment). • For white matter hyperintensities of presumed vascular
• Use of several scanners. origin: specify whether deep grey matter and brainstem
• Change of scanners or change to scanner system during study. hyperintensities are included (and if so, refer to all
• MRI sequences, acquisition parameters (including as hyperintensities collectively); state the rating scale or
appropriate: repetition time, echo time, inversion time, volume measurement software used and observer
echo train length), acquisition and reconstruction matrices, reliability; specify whether the white matter hyperintensity
field-of-view, slice thickness including gaps and scanning volume was adjusted for intracranial or brain volume and
plane, details of selected options (tailored excitation pulses, how this was done; state whether lacunes were included in
parallel imaging, flow compensation, preparation pulses, white matter hyperintensities or measured separately and
etc), and total acquisition time. If a work-in-progress whether acute lesions were masked.
package is used, provide as much information as possible. • For perivascular spaces: separate perivascular spaces of the
basal ganglia and white matter; describe how qualitative
Image analysis and postprocessing
aspects (number, location, size, etc) are defined; state the
• Use and qualification of a central analysis facility, or training
observer reliability of the rating scale.
procedure across several analysis centres.
• For cerebral microbleeds: specify number and distribution
• Whether analyses were done blinded to initial presentation
divided into lobar, deep, and infratentorial (brainstem and
or to other data (should be specified) that might affect
cerebellum); provide full details of image acquisition
interpretation.
parameters; specify application of standardised rating
• Details of qualitative visual rating and quantitative
scales.
computational methods, including the URL if available for
• For atrophy: specify the rating scale or method of volume
download or an appendix describing the method in detail.
measurement, whether corrected for intracranial volume,
• For visual rating scales: whether images were rated centrally
and method used to do this.
by one or a small number of readers; the raters’ background

www.thelancet.com/neurology Vol 12 August 2013 833

 Position Paper

Panel 3: Questions to be addressed in the future study of imaging small vessel disease

Recent small subcortical infarct • What is the clinical value of imaging of cerebral microbleeds to guide
• How can causally distinct subtypes (eg, embolic, parent, or branch artery treatment decisions (eg, thrombolysis and antithrombotic therapy)?
atheroma or intrinsic small vessel disease) be better differentiated with • How do results from different protocols for imaging and quantifying
the use of neuroimaging? microbleeds compare, and how can findings made using different
• What proportion of infarcts cavitate, become a white matter protocols be integrated into a common result.
hyperintensity, or disappear on conventional MRI, and what are the
Brain atrophy
factors that affect infarct conversion between the different states?
• To what extent do vascular lesions related to small vessel disease affect
Lacunes of presumed vascular origin measures of brain atrophy (eg, cortical vs subcortical regions, grey vs
• How can methods for distinguishing between lacunes and perivascular white matter?)
spaces be improved? • To what extent does brain atrophy (cortical and subcortical, and grey and
• How do secondary degenerative changes near lacunes affect their size and white matter) affect volumentric measures of vascular lesions?
shape in the long term? • How do vascular lesions mediate secondary brain atrophy?
• How important is the presence of a hyperintense rim when • What are the clinical consequences of secondary brain atrophy that was
differentiating between lacunes and other small cystic structures—eg, mediated by vascular lesions?
perivascular spaces?
Other vascular lesions
White matter hyperintensity of presumed vascular origin • How frequently do microinfarcts occur in neurodegenerative diseases?
• How different are the mechanisms and clinical consequences of What are the clinical consequences of microinfarcts? What are the
periventricular white matter hyperintensities and deep white matter imaging signatures of acute microinfarcts during clinical scanning at 1·5 T
hyperintensities? To address this question, clear definitions and and 3·0 T? How frequent are microinfarcts in Alzheimer’s disease?
standardised protocols are needed to distinguish periventricular white • How frequently does acute sulcal subarachnoid haemorrhage result in
matter hyperintensities from deep white matter hyperintensities. siderosis? To what extent does superficial siderosis affect cortical neuronal
• How can newer imaging techniques, such as measurement of T1 water function? What are the clinical consequences of superficial siderosis
content, help to improve characterisation of white matter hyperintensities (including the frequency of subsequent symptomatic subarachnoid
and determine the effect of white matter hyperintensities on connected haemorrhage or intracerebral haemorrhage)?
brain regions (eg, cortex), and on clinical symptoms?
General
• Is there pathological or epidemiological justification for distinguishing
• How can direct imaging of small intracerebral arteries be improved?
between hyperintensities in grey matter and those in white matter?
• How can subtle changes of vascular origin be differentiated from changes
Perivascular space mediated by other pathologies (eg, neurodegeneration), and what are the
• What are the mechanisms that cause multiple occurrences of perivascular exact histopathological substrates?
spaces? • To what extent do lesions related to small vessel disease cause secondary
• What is the link between perivascular space diameter and risk factors and neurodegeneration, and what are the mechanisms for this process?
potential clinical consequences? Can a perivascular space diameter be • How do vascular lesions interact with neurodegenerative pathological
identified so that perivascular spaces of greater diameter can be thought changes to cause cognitive decline and other clinical manifestations of
of as pathologically enlarged? neurodegeneration, particularly physical disability, gait disturbance, and
• How are perivascular spaces linked to small vessel disease-related vascular depression?
and parenchymal lesions? • Is there any suitable imaging scale that integrates many imaging
• How are perivascular spaces linked to brain atrophy and findings for small vessel disease and how would this scale be used in
neurodegenerative pathological changes? research and clinical practice (eg, for risk prediction and treatment
• How can the protocols for detecting and quantifying perivascular spaces stratification)?
be improved? • How can the imaging of vascular manifestations be used to improve the
• What are the clinical consequences of perivascular spaces? selection of individuals in future clinical trials and when measuring
• What is the clinical value of assessing perivascular spaces? outcomes?
• How can the protocols for measuring blood–brain barrier integrity be
Cerebral microbleed
improved, and how do changes related to small vessel disease and
• How do different neurodegenerative or vascular disorders affect the
neurodegenerative pathological changes interact to affect blood–brain
pathology and spatial distribution of microbleeds and their appearance on
barrier integrity?
imaging?
• What is the link between changes in cerebral blood flow and volume and
• How reliable, sensitive, or specific is the distribution of cerebral
ischaemic brain pathological changes?
microbleeds as a marker for cerebral amyloid angiopathy or hypertension?
• What is the predictive value of imaging cerebral microbleeds when
investigating cognitive and functional decline?

834 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

are advances in small-artery imaging with ultra-high-field genetic association studies. We suggest that our reporting
strength MRI (>3·0 T), diffusion tensor imaging of standards could be a useful guide to authors, peer
detailed structural connectivity, magnetisation transfer reviewers, and editors, and might raise the methodological
assessment of white matter myelination, imaging of quality of neuroimaging studies of SVD as have analogous
blood–brain barrier permeability and of vascular reactivity, reporting standards (eg, the CONSORT guidelines)
perfusion imaging, imaging of retinal vessels, quantitative increased the quality of other types of research. Finally,
susceptibility imaging, and others. The advantages of although the primary purpose of our standards was for
multimodal imaging should be used to advance research, we hope that these standards will have a positive
understanding of pathophysiology. For example, effect on clinical diagnosis and the reporting of SVD
combining diffusion tensor imaging and atrophy allows lesions in routine clinical practice.
the study of the effects of vascular lesions on distantly These standards were developed on the basis of
connected brain regions.108 consensus in a large group of experts using rigorous
Interest is increasing in the clinical relevance of methods. However, we acknowledge that in some parts
microinfarcts, another small infarct subtype. Microinfarcts of this rapidly evolving specialty, our recommendations
are very small ischaemic lesions recorded mostly in the are based on small studies, unvalidated findings, or
cortex during autopsy in older people, but are occasionally expert opinion without a strong body of supporting
seen on high-field MRI.109,110 Their true vascular pathological evidence. More work is essential to validate these
relation to SVD is still unknown. Although yet to be standards. In particular, the lack of radiological–
confirmed, some very small cortical acute lesions seen on pathological association studies is a key limitation to
diffusion-weighted imaging might actually be micro- our understanding of the neuroimaging findings in
infarcts.109 Microinfarcts seem very clinically relevant, so SVD. Additionally, best standards for neuroimaging of
neuroimaging methods to measure them or a validated SVD will change over time and, therefore, periodic
close surrogate are needed. Larger cortical microinfarcts revision of these standards will be needed.
have been visualised on 7·0 T MRI, and these lesions were Contributors
sometimes visible on conventional 3·0 T MRI.111 All authors contributed to the Position Paper through attendance at the
We now need comprehensive studies that account for two workshops, participation in specific workgroups, writing and editing
the report, and providing approval for final submission. JMW, MDi, and
both vascular and neurodegenerative pathological EES coordinated the overall process, obtained funding, organised the
changes in patients observed for longer and from a two workshops, coordinated the report preparation, and finalised the
younger age than have been done so far (panel 3). These paper for submission. PBG, CCh, BN, MvB, and CS were external
studies should incorporate some of the following: advisers who attended the second workshop, read the draft standards in
detail, provided in-depth critique, and approved the final paper for
multimodal MRI, amyloid imaging with PET, detailed submission. EES, FD, MDi, JMW, AV, JTOB, SEB, and RIL did the
clinical testing, and biomarkers from serum and CSF to literature searches. Workgroup participants were as follows: recent small
elucidate the crosstalk between vascular and subcortical infarcts (JMW [chair], RvO, VH, BCMS, VM, LP, and FD),
neurodegenerative pathological changes, and how these lacunes of presumed vascular origin (RIL [chair], HC, MB, CB, VM, and
ORB), white matter hyperintensities of presumed vascular origin (CCo
interact to cause clinical symptoms, particularly cognitive [chair], SEB, CDC, F-EdL, FF, JTOB, and LP), perivascular space (FF
decline, gait impairment, physical disability, and [chair], VH, AV, GJB, RvO, and FB), cerebral microbleeds (JTOB [chair],
depression. Researchers should also investigate the DW, MB, CCo, RIL, and RF), brain atrophy (GJB [chair], NF, IK, SG, ST,
connection between imaging and pathological SEB, and FB), image acquisition and analysis (RF, OS, SEB, MDu, and
JMW), clinical relevance (ORB and LP), other lesions (EES [chair], DW,
changes.30,52 Additionally, carefully conducted imaging and AV), and future directions (MDi and MDu).
studies with short-term (eg, monthly) observational
Conflicts of interest
periods are needed to study the progression of imaging GJB consults for and receives research support from Boehringer
features of SVD and their effects on brain tissue loss. Ingelheim. FF serves on scientific advisory boards for Bayer Schering,
Biogen Idec, Genzyme, Merck Serono, Pfizer, Novartis, and Teva
Pharmaceutical Industries, and as scientific adviser for Perceptive
Conclusions Informatics, he serves on the editorial boards of Cerebrovascular Diseases,
Our neuroimaging standards for SVD are intended to Multiple Sclerosis, the Polish Journal of Neurology and Neurosurgery, Stroke,
harmonise current disparate terminology and analysis and the Swiss Archives of Neurology and Psychiatry, and has received
methods. We encourage other researchers to use these speaker honoraria and support from Biogen Idec, Bayer Schering, Merck
Serono, Novartis, Sanofi-Aventis, and Teva Pharmaceutical Industries.
standards. They could provide a useful introduction to the
RF receives operating research funding from the Canadian Institutes of
principles of neuroimaging in SVD, which could help Health Research, the Alberta and Pfizer Translational Research Program,
investigators to incorporate measurement of SVD into the Hotchkiss Brain Institute, and the University of Calgary to support
studies of neurodegenerative diseases or the pathological work on imaging SVD. JTOB has been a consultant for GE Healthcare,
Lilly, Bayer Healthcare, and TauRx, and has received speakers’ honoraria
basis of ageing. These standards should allow cross-study
from Pfizer, GE Healthcare, Eisai, Shire, Lundbeck, Lilly, and Novartis.
comparisons of findings, accelerating the translation of SEB has had research funds to the Cognitive Neurology Research and
new findings into practice. Additionally, standardised Stroke Research Units from Pfizer (Wyeth Research), GlaxoSmithKline,
approaches could enable formal between-study meta- Lundbeck, Roche, and Novartis. She has received speakers’ honoraria for
continuing medical education from Pfizer, Eisai, and Novartis, and
analyses, which are increasingly needed to obtain the honoraria for ad-hoc consulting from Pfizer, Novartis, GlaxoSmithKline,
sample sizes necessary for some types of research such as

www.thelancet.com/neurology Vol 12 August 2013 835

 Position Paper

Roche, Elan, and Bristol-Myers Squibb. HC has received honoraria for 13 Editorial. A united approach to vascular disease and
consultancy from Servier, Lundbeck, and Janssen Research and neurodegeneration. Lancet Neurol 2012; 11: 293.
Development. F-EdL is supported by a Vidi grant (016·126·351) from the 14 Kang DW, Han MK, Kim HJ, et al. New ischemic lesions
Netherlands Organisation for Health Research and Development. LP is a coexisting with acute intracerebral hemorrhage. Neurology 2012;
member of the editorial boards of Acta Neurologica Scandinavica, 79: 848–55.
International Journal of Alzheimer Disease, and Cerebrovascular Diseases 15 Chowdhury D, Wardlaw JM, Dennis MS. Are multiple acute small
and editor of the vascular cognitive impairment section of Stroke. ST subcortical infarctions caused by embolic mechanisms?
served as a member of the advisory board for Lilly Deutschland. MDi is a J Neurol Neurosurg Psychiatry 2004; 75: 1416–20.
consultant for Lilly Deutschland and Boehringer Ingelheim, Biologische 16 Doubal FN, Dennis MS, Wardlaw JM. Characteristics of patients
Heilmittel Heel, Bristol-Myers Squibb, and Ever Neuro Pharma. He has with minor ischaemic strokes and negative MRI: a cross sectional
received honoraria from Bayer Vital, Boehringer Ingelheim Pharma study. J Neurol Neurosurg Psychiatry 2011; 82: 540–42.
Biologische Heilmittel Heel, Bristol-Myers Squibb, Lundbeck, 17 Potter GM, Doubal FN, Jackson CA, et al. Counting cavitating
Sanofi-Aventis Deutschland, Shire Deutschland, German Centre for lacunes underestimates the burden of lacunar infarction.
Stroke 2010; 41: 267–72.
Neurodegenerative Diseases (DZNE), Georg Thieme Verlag, UpToDate,
and W Kohlhammer. He has received Principal Investigator and 18 Moreau F, Patel S, Lauzon ML, et al. Cavitation after acute
symptomatic lacunar stroke depends on time, location, and MRI
Research Funding (industry) from Bayer Vital, Eisai Medical Research,
sequence. Stroke 2012; 43: 1837–42.
Eisai, Essex Pharma, Ferrer Internacional, and ICON Clinical Research.
19 Koch S, McClendon MS, Bhatia R. Imaging evolution of lacunar
All other authors declare that they have no conflicts of interest.
stroke–leukoariosis or lacune? Neurology 2011; 77: 1091–95.
Acknowledgments 20 Donnan GA, Bladin PF, Berkovic SF, Longley WA, Saling MM.
The following institutions provided funding: the Centres of Excellence in The stroke syndrome of striatocapsular infarction.
Neurodegeneration (COEN) concordat (UK Medical Research Council, Brain 1991; 114: 51–70.
the German Centre for Neurodegenerative Disease [DZNE], and the 21 Fisher CM. Lacunar infarcts—a review. Cerebrovasc Dis
Canadian Institutes of Health Research [CIHR]) reference COEN017; 1991; 1: 311–20.
Royal Society of Edinburgh; Scottish Imaging Network, a Platform for 22 Vermeer SE, Longstreth WT Jr, Koudstaal PJ. Silent brain infarcts:
Scientific Excellence (SINAPSE) Collaboration; UK Cross-Council Centre a systematic review. Lancet Neurol 2007; 6: 611–19.
for Cognitive Ageing and Cognitive Epidemiology, Edinburgh, UK; 23 Santos M, Gold G, Kovari E, et al. Differential impact of lacunes and
Carl Friedrich von Siemens Foundation; Vascular Dementia Research microvascular lesions on poststroke depression.
Foundation; NIHR Biomedical Research Unit in Dementia awarded to Stroke 2009; 40: 3557–62.
Cambridge University Hospitals NHS Trust and the University of 24 Choi P, Ren M, Phan TG, et al. Silent infarcts and cerebral
Cambridge, Cambridge, UK; NIHR Biomedical Research Unit in microbleeds modify the associations of white matter lesions with
Dementia awarded to UCL Hospitals NHS Trust and UCL, London UK; gait and postural stability: population-based study.
Stroke 2012; 43: 1505–10.
FP6 ERA-NET NEURON grant (01 EW1207); and the Canadian Stroke
Network. We thank M Henderson, K Shuler, L Thomas, I Wittko, and 25 Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA,
Markesbery WR. Brain infarction and the clinical expression of
Anna Charlton for administrative support.
Alzheimer disease. The Nun Study. JAMA 1997; 277: 813–17.
References 26 Franke CL, van Swieten JC, van Gijn J. Residual lesions on
1 Norrving B. Lacunar infarcts: no black holes in the brain are computed tomography after intracerebral hemorrhage.
benign. Pract Neurol 2008; 8: 222–28. Stroke 1991; 22: 1530–33.
2 Pantoni L. Cerebral small vessel disease: from pathogenesis and 27 Vermeer SE, den Heijer T, Koudstaal PJ, Oudkerk M, Hofman A,
clinical characteristics to therapeutic challenges. Breteler MM. Incidence and risk factors of silent brain infarcts in
Lancet Neurol 2010; 9: 689–701. the population-based Rotterdam Scan Study.
3 Kivipelto M, Helkala EL, Laakso M, et al. Midlife vascular risk Stroke 2003; 34: 392–96.
factors and Alzheimer’s disease in later life: longitudinal, 28 Longstreth WT Jr, Bernick C, Manolio TA, Bryan N, Jungreis CA,
population based study. BMJ 2001; 322: 1447–51. Price TR. Lacunar infarcts defined by magnetic resonance imaging
4 Dichgans M, Zietemann V. Prevention of vascular cognitive of 3660 elderly people. The Cardiovascular Health Study.
impairment. Stroke 2012; 43: 3137–46. Arch Neurol 1998; 55: 1217–25.
5 Brayne C, Richardson K, Matthews FE, et al. Neuropathological 29 Debette S, Markus HS. The clinical importance of white matter
correlates of dementia in over-80-year-old brain donors from the hyperintensities on brain magnetic resonance imaging: systematic
population-based Cambridge City over-75s cohort (CC75C) study. review and meta-analysis. BMJ 2010; 341: c3666.
J Alzheimers Dis 2009; 18: 645–58. 30 Gouw AA, Seewann A, van der Flier WM, et al. Heterogeneity of
6 Matthews FE, Brayne C, Lowe J, McKeith I, Wharton SB, Ince P. small vessel disease: a systematic review of MRI and
Epidemiological pathology of dementia: attributable-risks at death histopathology correlations. J Neurol Neurosurg Psychiatry
in the Medical Research Council Cognitive Function and Ageing 2011; 82: 126–35.
Study. PLoS Med 2009; 6: e1000180. 31 de Laat KF, Tuladhar AM, van Norden AGW, Norris DG,
7 Neuropathology Group of the Medical Research Council. Cognitive Zwiers MP, de Leeuw F-E. Loss of white matter integrity is
function and ageing study (MRC CFAS). Lancet 2001; 357: 169–75. associated with gait disorders in cerebral small vessel disease.
8 Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain Brain 2011; 134: 73–83.
pathologies account for most dementia cases in community- 32 Windham BG, Griswold ME, Shibata D, Penman A, Catellier DJ,
dwelling older persons. Neurology 2007; 69: 2197–204. Mosley TH Jr. Covert neurological symptoms associated with silent
9 Potter GM, Marlborough FJ, Wardlaw JM. Wide variation in infarcts from midlife to older age: the Atherosclerosis Risk in
definition, detection, and description of lacunar lesions on imaging. Communities study. Stroke 2012; 43: 1218–23.
Stroke 2010; 42: 359–66. 33 Saini M, Ikram K, Hilal S, Qiu A, Venketasubramanian N, Chen C.
10 Zhu YC, Dufouil C, Tzourio C, Chabriat H. Silent brain infarcts: Silent stroke: not listened to rather than silent. Stroke 2012;
a review of MRI diagnostic criteria. Stroke 2011; 42: 1140–45. 3: 3102–04.
11 Hachinski V, Iadecola C, Petersen RC, et al. National Institute of 34 Haley AP, Hoth KF, Gunstad J, et al. Subjective cognitive
Neurological Disorders and Stroke–Canadian Stroke Network complaints relate to white matter hyperintensities and future
vascular cognitive impairment harmonization standards. cognitive decline in patients with cardiovascular disease.
Stroke 2006; 37: 2220–41. Am J Geriatr Psychiatry 2009; 17: 976–85.
12 Gorelick PB, Scuteri A, Black SE, et al. Vascular contributions to 35 Inzitari D, Pracucci G, Poggesi A, et al. Changes in white matter as
cognitive impairment and dementia: a statement for healthcare determinant of global functional decline in older independent
professionals from the American Heart Association/American outpatients: three year follow-up of LADIS (leukoaraiosis and
Stroke Association. Stroke 2011; 42: 2672–713. disability) study cohort. BMJ 2009; 339: 279–82.

836 www.thelancet.com/neurology Vol 12 August 2013

 Position Paper

36 Hachinski VC, Potter P, Merskey H. Leukoaraiosis. 57 Cordonnier C, Al-Shahi Salman R, Wardlaw J. Spontaneous brain
Arch Neurol 1987; 44: 21–23. microbleeds: systematic review, subgroup analyses and standards
37 Pollock H, Hutchings M, Weller RO, Zhang E-T. Perivascular for study design and reporting. Brain 2007; 130: 1988–2003.
spaces in the basal gangli of the human brain: their relationship to 58 Vernooij MW, van der Lugt A, Ikram MA, et al. Prevalence and risk
lacunes. J Anat 1997; 191: 337–46. factors of cerebral microbleeds: the Rotterdam Scan Study.
38 Groeschel S, Chong WK, Surtees R, Hanefeld F. Virchow-Robin Neurology 2008; 70: 1208–14.
spaces on magnetic resonance images: normative data, their 59 Knudsen KA, Rosand J, Karluk D, Greenberg SM. Clinical diagnosis
dilatation, and a review of the literature. of cerebral amyloid angiopathy: validation of the Boston criteria.
Neuroradiology 2006; 48: 745–54. Neurology 2001; 56: 537–39.
39 Doubal FN, MacLullich AM, Ferguson KJ, Dennis MS, Wardlaw JM. 60 Greenberg SM, Vernooij MW, Cordonnier C, et al. Cerebral
Enlarged perivascular spaces on MRI are a feature of cerebral small microbleeds: a guide to detection and interpretation. Lancet Neurol
vessel disease. Stroke 2010; 41: 450–54. 2009; 8: 165–74.
40 Kwee RM, Kwee TC. Virchow-Robin spaces at MR imaging. 61 Sperling RA, Jack CR Jr, Black SE, et al. Amyloid-related imaging
Radiographics 2007; 27: 1071–86. abnormalities in amyloid-modifying therapeutic trials:
41 Zhu YC, Tzourio C, Soumare A, Mazoyer B, Dufouil C, recommendations from the Alzheimer’s Association Research
Chabriat H. Severity of dilated Virchow-Robin spaces is Roundtable Workgroup. Alzheimers Dement 2011; 7: 367–85.
associated with age, blood pressure, and MRI markers of small 62 Lammie GA, Lindley R, Keir S, Wiggam MI. Stress-related primary
vessel disease: a population-based study. intracerebral hemorrhage: autopsy clues to underlying mechanism.
Stroke 2010; 41: 2483–90. Stroke 2000; 31: 1426–48.
42 MacLullich AM, Wardlaw JM, Ferguson KJ, Starr JM, Seckl JR, 63 Broderick J, Connolly S, Feldmann E, et al. Guidelines for the
Deary IJ. Enlarged perivascular spaces are associated with cognitive management of spontaneous intracerebral hemorrhage in adults:
function in healthy elderly men. 2007 update: a guideline from the American Heart Association/
J Neurol Neurosurg Psychiatry 2004; 75: 1519–23. American Stroke Association Stroke Council, High Blood Pressure
43 Poirier J, Derouesne C. Cerebral lacunae. A proposed new Research Council, and the Quality of Care and Outcomes in Research
classification. Clin Neuropathol 1984; 3: 266. Interdisciplinary Working Group. Stroke 2007; 38: 2001–202.
44 Heier LA, Bauer CJ, Schwartz L, Zimmerman RD, Morgello S, 64 Qureshi AI, Tuhrim S, Broderick JP, Batjer HH, Hondo H,
Deck MDF. Large Virchow-Robin spaces: MR-clinical correlation. Hanley DF. Spontaneous intracerebral hemorrhage.
AJNR Am J Neuroradiol 1989; 10: 929–36. N Engl J Med 2001; 344: 1450–60.
45 Patankar TF, Mitra D, Varma A, Snowden J, Neary D, Jackson A. 65 Jackson CA, Sudlow CL. Is hypertension a more frequent risk factor
Dilatation of the Virchow-Robin space is a sensitive indicator of for deep than for lobar supratentorial intracerebral haemorrhage?
cerebral microvascular disease: study in elderly patients with J Neurol Neurosurg Psychiatry 2006; 77: 1244–52.
dementia. AJNR Am J Neuroradiol 2005; 26: 1512–20. 66 Bhattathiri PS, Gregson B, Prasad KSM, et al. Reliability assessment
46 Rouhl RPW, van Oostenbrugge RJ, Knottnerus ILH, Staals JEA, of computerized tomography scanning measurements in
Lodder J. Virchow-Robin spaces relate to cerebral small vessel intracerebral hematoma. Neurosurg Focus 2003; 15: 1–5.
disease severity. J Neurol 2008; 255: 692–96. 67 Linn J, Halpin A, Demaerel P, et al. Prevalence of superficial
47 Zhu YC, Dufouil C, Mazoyer B, et al. Frequency and location of siderosis in patients with cerebral amyloid angiopathy. Neurology
dilated Virchow-Robin spaces in elderly people: a population- 2010; 74: 1346–50.
based 3D MR imaging study. AJNR Am J Neuroradiol 2011; 68 Kumar N, Cohen-Gadol AA, Wright RA, Miller GM, Piepgras DG,
32: 709–13. Ahlskog JE. Superficial siderosis. Neurology 2006; 66: 1144–52.
48 Chen W, Song X, Zhang Y. Assessment of the Virchow-Robin 69 Jack CR Jr. Alliance for aging research AD biomarkers work group:
Spaces in Alzheimer disease, mild cognitive impairment, and structural MRI. Neurobiol Aging 2011; 32: S48–57.
normal aging, using high-field MR imaging. 70 Walters RJ, Fox NC, Crum WR, Taube D, Thomas DJ.
AJNR Am J Neuroradiol 2011; 32: 1490–95. Haemodialysis and cerebral oedema. Nephron 2001; 87: 143–47.
49 Braffman BH, Zimmerman RA, Trojanowski JQ, Gonatas NK, 71 Duning T, Kloska S, Steinstrater O, Kugel H, Heindel W, Knecht S.
Hickey WF, Schlaepfer WW. Brain MR: pathologic correlation with Dehydration confounds the assessment of brain atrophy. Neurology
gross and histopathology. 2. Hyperintense white-matter foci in the 2005; 64: 548–50.
elderly. AJR Am J Roentgenol 1988; 151: 559–66. 72 Bobinski M, de Leon MJ, Wegiel J, et al. The histological validation of
50 Bokura H, Kobayashi S, Yamaguchi S. Distinguishing silent lacunar post mortem magnetic resonance imaging-determined hippocampal
infarction from enlarged Virchow-Robin spaces: a magnetic volume in Alzheimer’s disease. Neuroscience 2000; 95: 721–25.
resonance imaging and pathological study. J Neurol 1998; 73 Jagust WJ, Zheng L, Harvey DJ, et al. Neuropathological basis of
245: 116–22. magnetic resonance images in aging and dementia. Ann Neurol
51 Awad IA, Johnson PC, Spetzler RF, Hodak JA. Incidental 2008; 63: 72–80.
subcortical lesions identified on magnetic resonance imaging in the 74 Black S, Gao F, Bilbao J. Understanding white matter disease.
elderly. II Postmortem pathological correlations. Stroke 1986; Imaging–pathological correlations in vascular cognitive
17: 1090–97. impairment. Stroke 2009; 40: S48–52.
52 Shoamanesh A, Kwok CS, Benavente O. Cerebral microbleeds: 75 Appelman AP, Exalto LG, van der Graaf Y, Biessels GJ, Mali WP,
histopathological correlation of neuroimaging. Cerebrovasc Dis 2011; Geerlings MI. White matter lesions and brain atrophy: more than
32: 528–34. shared risk factors? A systematic review. Cerebrovasc Dis 2009;
53 De Reuck J, Auger F, Cordonnier C, et al. Comparison of 7·0-T T*- 28: 227–42.
magnetic resonance imaging of cerebral bleeds in post-mortem 76 Aribisala BS, Valdes Hernandez MC, Royle NA, et al. Brain atrophy
brain sections of Alzheimer patients with their neuropathological associations with white matter lesions in the ageing brain:
correlates. Cerebrovasc Dis 2011; 31: 511–17. the Lothian Birth Cohort 1936. Eur Radiol 2013; 23: 1084–92.
54 Fazekas F, Kleinert R, Roob G, et al. Histopathologic analysis of foci 77 Mungas D, Jagust WJ, Reed BR, et al. MRI predictors of cognition
of signal loss on gradient-echo T2*-weighted MR images in patients in subcortical ischemic vascular disease and Alzheimer’s disease.
with spontaneous intracerebral hemorrhage: evidence of Neurology 2001; 57: 2229–35.
microangiopathy-related microbleeds. AJNR Am J Neuroradiol
78 Levy-Cooperman N, Ramirez J, Lobaugh NJ, Black SE. Misclassified
1999; 20: 637–42.
tissue volumes in Alzheimer disease patients with white matter
55 Dichgans M, Holtmannspotter M, Herzog J, Peters N, hyperintensities: importance of lesion segmentation procedures for
Bergmann M, Yousry TA. Cerebral microbleeds in CADASIL: volumetric analysis. Stroke 2008; 39: 1134–41.
a gradient-echo magnetic resonance imaging and autopsy study.
79 Carmichael O, Schwarz C, Drucker D, et al. Longitudinal changes
Stroke 2002; 33: 67–71.
in white matter disease and cognition in the first year of the
56 Tatsumi S, Shinohara M, Yamamoto T. Direct comparison of Alzheimer disease neuroimaging initiative. Arch Neurol 2010;
histology of microbleeds with postmortem MR images: a case 67: 1370–78.
report. Cerebrovasc Dis 2008; 26: 142–46.

www.thelancet.com/neurology Vol 12 August 2013 837

 Position Paper

80 Wardlaw JM, Brindle W, Casado AM, et al. A systematic review of 96 Valdes Hernandez MC, Ferguson KJ, Chappell FM, Wardlaw JM.
the utility of 1·5 versus 3 Tesla magnetic resonance brain imaging New multispectral MRI data fusion technique for white matter
in clinical practice and research. Eur Radiol 2012; 22: 2295–303. lesion segmentation: method and comparison with thresholding in
81 Potter G, Doubal F, Jackson C, Sudlow C, Dennis M, Wardlaw J. FLAIR images. Eur Radiol 2010; 20: 1684–91.
Associations of clinical stroke misclassification (‘clinical–imaging 97 Firbank MJ, Minett T, O’Brien JT. Changes in DWI and MRS
dissociation’) in acute ischemic stroke. Cerebrovasc Dis 2010; associated with white matter hyperintensities in elderly subjects.
29: 395–402. Neurology 2003; 61: 950–54.
82 Valdes Hernandez MC, Morris Z, Dickie DA, et al. Close correlation 98 Barnes SR, Haacke EM, Ayaz M, Boikov AS, Kirsch W, Kido D.
between quantitative and qualitative assessments of white matter Semiautomated detection of cerebral microbleeds in magnetic
lesions. Neuroepidemiology 2012; 40: 13–22. resonance images. Magn Reson Imaging 2011; 29: 844–52.
83 Gao FQ, Swartz RH, Scheltens P, et al. Complexity of MRI white 99 Seghier ML, Kolanko MA, Leff AP, Jager HR, Gregoire SM,
matter hyperintensity assessments in relation to cognition in aging Werring DJ. Microbleed detection using automated segmentation
and dementia from the Sunnybrook Dementia Study. (MIDAS): a new method applicable to standard clinical MR images.
J Alzheimers Dis 2011; 26 (suppl 3): 379–88. PLoS One 2011; 6: e17547.
84 Wang X, Valdes Hernandez MC, Doubal F, Chappell FM, 100 Cordonnier C, Potter GM, Jackson CA, et al. Improving interrater
Wardlaw JM. How much do focal infarcts distort white matter agreement about brain microbleeds. Development of the brain
lesions and global cerebral atrophy measures? Cerebrovasc Dis observer microbleed scales (BOMBS). Stroke 2009; 49: 94–99.
2012; 34: 336–42. 101 Gregoire SM, Chaudhary UJ, Brown MM, et al. The microbleed
85 Wong KS, Gao S, Chan YL, et al. Mechanisms of acute cerebral anatomical rating scale (MARS): reliability of a tool to map brain
infarctions in patients with middle cerebral artery stenosis: a microbleeds. Neurology 2009; 73: 1759–66.
diffusion-weighted imaging and microemboli monitoring study. 102 Scheltens P, Pasquier F, Weerts JG, Barkhof F, Leys D. Qualitative
Ann Neurol 2002; 52: 74–81. assessment of cerebral atrophy on MRI: inter- and intra-observer
86 Seifert T, Enzinger C, Storch MK, Pichler G, Niederkorn K, reproducibility in dementia and normal aging. Eur Neurol 1997;
Fazekas F. Acute small subcortical infarctions on diffusion 37: 95–99.
weighted MRI: clinical presentation and aetiology. 103 Farrell C, Chappell F, Armitage PA, et al. Development and initial
J Neurol Neurosurg Psychiatry 2005; 76: 1520–24. testing of normal reference MR images for the brain at ages 65–70
87 Longstreth WT Jr, Manolio TA, Arnold A, et al. Clinical correlates of and 75–80 years. Eur Radiol 2008; 19: 177–83.
white matter findings on cranial magnetic resonance imaging 104 Ashburner J, Csernansky JG, Davatzikos C, Fox NC, Frisoni GB,
of 3301 elderly people. The Cardiovascular Health Study. Stroke Thompson PM. Computer-assisted imaging to assess brain
1996; 27: 1274–82. structure in healthy and diseased brains. Lancet Neurol 2003;
88 Wahlund LO, Barkhof F, Fazekas F, et al. A new rating scale for 2: 79–88.
age-related white matter changes applicable to MRI and CT. 105 O’Sullivan M, Jouvent E, Saemann PG, et al. Measurement of brain
Stroke 2001; 32: 1318–22. atrophy in subcortical vascular disease: a comparison of different
89 Bocti C, Swartz RH, Gao FQ, Sahlas DJ, Behl P, Black SE. A new approaches and the impact of ischaemic lesions. Neuroimage 2008;
visual rating scale to assess strategic white matter hyperintensities 43: 312–20.
within cholinergic pathways in dementia. Stroke 2005; 36: 2126–31. 106 de Bresser J, Portegies MP, Leemans A, Biessels GJ, Kappelle LJ,
90 Prins ND, van Straaten ECW, van Dijk EJ, et al. Measuring Viergever MA. A comparison of MR based segmentation methods
progression of cerebral white matter lesions on MRI. Visual rating for measuring brain atrophy progression. Neuroimage 2011;
and volumetrics. Neurology 2004; 62: 1533–39. 54: 760–68.
91 Kapeller P, Barber R, Vermeulen RJ, et al. Visual rating of age- 107 Frisoni GB, Jack CR. Harmonization of magnetic resonance-based
related white matter changes on magnetic resonance imaging: scale manual hippocampal segmentation: a mandatory step for wide
comparison, interrater agreement, and correlations with clinical use. Alzheimers Dement 2011; 7: 171–74.
quantitative measurements. Stroke 2003; 34: 441–45. 108 Duering M, Righart R, Csandi E, et al. Incident subcortical infarcts
92 Fazekas F, Barkhof F, Wahlund LO, et al. CT and MRI rating of induce focal thinning in connected cortical regions. Neurology 2012;
white matter lesions. Cerebrovasc Dis 2003; 13: 31–36. 79: 2025–28 .
93 Ramirez J, Gibson E, Quddus A, et al. Lesion Explorer: a 109 Smith EE, Schneider JA, Wardlaw JM, Greenberg SM. Cerebral
comprehensive segmentation and parcellation package to obtain microinfarcts: the invisible lesions. Lancet Neurol 2012; 11: 272–82.
regional volumetrics for subcortical hyperintensities and 110 Brundel M, de Bresser J, van Dillen JJ, Kappelle LJ, Biessels GJ.
intracranial tissue. Neuroimage 2011; 54: 963–73. Cerebral microinfarcts: a systematic review of neuropathological
94 Jeerakathil T, Wolf PA, Beiser A, et al. Stroke risk profile predicts studies. J Cereb Blood Flow Metab 2012; 32: 425–36.
white matter hyperintensity volume: the Framingham Study. 111 van Veluw SJ, Zwanenburg JJ, Engelen-Lee J, et al. In vivo detection
Stroke 2004; 35: 1857–61. of cerebral cortical microinfarcts with high-resolution 7T MRI.
95 van den Heuvel DMJ, ten Dam VH, de Craen AJ, et al. Measuring J Cereb Blood Flow Metab 2012; 33: 322–29.
longitudinal white matter changes: comparison of a visual rating
scale with a volumetric measurement. AJNR Am J Neuroradiol
2006; 27: 875–78.

838 www.thelancet.com/neurology Vol 12 August 2013