Journal of Alzheimer’s Disease 42 (2014) S365–S374DOI 10.3233/JAD-141803IOS Press
S365
Review
Utility of Transcranial Ultrasound inPredicting Alzheimer’s Disease Risk
Ales Tomek∗, Barbora Urbanova and Jakub HortDepartment of Neurology, 2nd Faculty of Medicine, Charles University in Prague and Motol University Hospital,Prague, Czech Republic
Accepted 1 October 2014
Abstract. Alzheimer’s disease (AD) is a progressive, neurodegenerative disease characterized by an increasing incidence. Oneof the pathologic processes that underlie this disorder is impairment of brain microvasculature. Transcranial ultrasound is anon-invasive examination of cerebral blood flow that can be employed as a simple and useful screening tool for assessing thevascular status of brain circulation in preclinical and clinical stages of AD. The objective of this review is to explore the utility ofusing a transcranial ultrasound to diagnose AD. With transcranial ultrasound, the most frequently studied parameters are cerebralblood flow velocities and pulsatility indices, cerebrovascular reserve capacity, and cerebral microembolization. On the basis ofcurrent knowledge, we recommend using as a transcranial Doppler sonography screening method of choice the assessment ofcerebrovascular reserve capacity with breath-holding test.
Giving the increasing fraction of older people inthe general population and increased life expectan-cies around the world, we are facing an epidemicwithout precedent: the unwanted side effects oflongevity—cognitive decline, disability, and depen-dency during aging. A hallmark of aging is theincreasing prevalence of Alzheimer’s disease (AD) andvascular dementia (VaD), with growing mortality fromcerebrovascular disease (CVD).
The vascular hypothesis of AD, first proposed byde la Torre 20 years ago, suggests that vascular riskfactors play a critical role in the development of cogni-tive decline and AD during aging [1]. Although CVD,VaD, and AD share identical risk factors, it still notclear whether CVD occurs as a parallel event or as atrigger or an accelerator factor for tau pathology in AD.The relationship between CVD and AD resembles the
∗Correspondence to: Ales Tomek, MD, PhD, FESO, Departmentof Neurology, 2nd Faculty of Medicine, Charles University in Pragueand Motol University Hospital, V Uvalu 84, Prague 5, 150 18, CzechRepublic. Tel.: +420 224 436 803; Fax: +420 224 436 875; E-mail:[email protected].
famous “chicken or the egg” causality dilemma andhas not been satisfactorily explained to date.
The most plausible hypothesis is that CVD couldaggravate amyloid and tau pathologies in patients withpreexisting genetic and lifestyle-related risk factorsfor AD thus leading to a faster clinical progressionof dementia. It was proposed that at least one roleof amyloid-� (A�) might be to serve as a part ofthe defense mechanism to seal small vessel leakage.This vascular repair system might be beneficial inyoung individuals. However, in the case of chronichypoperfusion, the same mechanism may lead to con-tinuous amyloid deposition that adversely promotesmicrovascular occlusion [2]. Another possible linkbetween hypoperfusion and the AD pathology couldbe that A� accumulation along pericapillary intersti-tial fluid drainage pathways could prevent the drainageof interstitial fluid to the cerebrospinal fluid, whichcould result in a decreased clearance of A� fromthe brain [3]. Furthermore, chronically hypoperfusedbrains of individuals with preexisting AD pathologieswith structural changes in brain microvasculature are
S366 A. Tomek et al. / Utility of Transcranial Ultrasound
then more susceptible to ischemic damage, which man-ifests itself as an increased incidence of white-matterlesions and microhaemorrhages [4]. Decreased cere-bral blood flow (CBF) observed in AD patients couldbe then related to microangiopathy aggravated by ADpathology or this decrease could be at least partiallycaused by decreased metabolic demand due to the lossof neurons in brain atrophy alone independently ofmicroangiopathy [5]. The decreased CBF in AD couldalso be at least in some individuals related to othercauses, i.e., cardiovascular disease with decreased car-diac output [6] or atherosclerotic disease of majorcerebral vessels as we discuss later.
Traditional neuroscience has first concentrated on asymptomatic treatment via the modulation of cholin-ergic or glutamatergic systems. Only later has sciencefocused on the degenerative part of the AD story withimmunotherapy and vaccination. Success in this fieldis still years away. A vascular approach was not a lead-ing research topic in AD treatments, although severalclinical trials are currently underway to examine thepossible benefit of vascular interventions to a courseof cognitive decline (e.g., PreDIVA, FINGER) [7].
Transcranial ultrasound may provide a usefulscreening tool for assessing the vascular status of braincirculation in the preclinical and clinical stages of AD.
The objective of this review is to explore the utilityof using transcranial ultrasound to diagnose AD,especially its preclinical or mild clinical stages of mildcognitive impairment (MCI). However, it is not clearwhether MCI, and specifically its amnestic subtype,represents a translational stage of evolving dementiaor is just an additional risk factor for AD [10]. Wealso discuss whether there is a special transcranialDoppler sonography (TCD) pattern of circulationimpairment in AD and MCI compared with VaD andhealthy control subjects.
TRANSCRANIAL ULTRASOUND
Aaslid et al. first described TCD for the non-invasiveexamination of CBF in 1982 [8]. The obtained Dopplersignal is assigned to a specific artery based on the expe-rience of the operator and indirect parameters only: thedepth of the sample volume, the position of the trans-ducer, and the direction of the flow (Fig. 1). Mistakesin the identification of individual vessels can occur inthe presence of anatomical variations or vessel pathol-ogy. Another setback of TCD is its inability to correctlymeasure velocities according to the angle between theinsonated vessel and the ultrasonic beam [9].
More than 30 years of technical developmentlater, transcranial color-coded duplex ultrasonography(TCCS) now makes high-resolution visualization ofcerebral arteries and brain parenchyma possible, evenwith simultaneous spatial visualization of the ultra-sound data on imported magnetic resonance images.The cerebral arteries can be exactly identified basedon their anatomic location with respect to each otherand surrounding brain structures. TCCS also offersthe possibility of visually correcting the insonationangle and thus obtaining more precise velocity mea-surements compared with conventional TCD (Fig. 2).Both methods are equal in pulsatility index (PI) andresistance index (RI) measurements. The visualizationof brain parenchyma in B-mode imaging on TCCS sys-tems is without clinical relevance in dementia patientscompared to its screening utility in Parkinson’s dis-ease patients [11]. TCD systems are usually equippedcontrary to TCSS with dedicated probe holders allow-ing for monitoring the flow for longer time periods.This feature makes TCD superior to TCCS for assess-ment of cerebrovascular reserve capacity (CVR) ormicroemboli monitoring.
The limitations of both TCD and TCCS, particularlyin the elderly population, are mainly related to an unfa-vorable acoustic bone window in 10–20% of patients.The inability to image intracranial vessels in thesecases can be overcome using echo contrast agents [9].
FLOW VELOCITIES ANDCEREBROVASCULAR RESISTANCE
The CBF curve recorded by transcranial ultrasounddirectly records two main flow velocities, peak sys-tolic velocity and end diastolic velocity. The otherparameters, mean flow velocity (MV), PI, and RI,can be derived from the flow curve. Velocity mea-surements can be obtained from all major intracranialvessels (branches of Willis circle)—anterior, middle,and posterior cerebral arteries, vertebral arteries, andthe basilar artery. PI and RI are measures of theblood flow curve and they reflect the resistance ofthe microvascular bed distal to the site of measure-ment. Although exact pathophysiological significanceof both indices is largely disputed in literature,they are influenced by several other hemodynamicfactors including cerebral perfusion pressure (dif-ference between mean arterial pressure and meanintracranial pressure) and blood rheology (especiallyhematocrit) [12–14]. This fact must be taken intoaccount when examining AD patients with other possi-ble pathology affecting the blood flow velocities (i.e.,
AU
THO
R C
OP
Y
A. Tomek et al. / Utility of Transcranial Ultrasound S367
Fig. 1. Example of transcranial Doppler ultrasound assessment of cerebrovascular reserve capacity in middle cerebral artery. Dedicated softwareenables direct visualization of flow velocities in a selected time frame with possibility of labeling the used vasoactive stimulus, e.g., breath-holdingor acetalozamide injection (lower frame).
Fig. 2. Example of transcranial color-coded duplex ultrasonography examination of middle cerebral artery (M2 segment). B-mode imaging ofanatomical landmarks with overlaid non directional color Doppler (power mode) with manually angle-corrected Doppler flow signal.
AU
THO
R C
OP
Y
S368 A. Tomek et al. / Utility of Transcranial Ultrasound
brain expansion, cardiac or respiratory insufficiency,or stenotic cerebrovascular disease). Both indices areusually derived automatically from the waveform bythe ultrasound device. The RI can be expressed as adifference between peak systolic velocity and end dias-tolic velocity divided by peak systolic velocity. The PIcan be expressed as a difference between peak systolicvelocity and end diastolic velocity divided by MFV.Normal values for RI are 0.49–0.63 and 0.6–1.1 for PI[15]. Both indices increase with age [16].
The type of transcranial ultrasound system (TCDor TCCS) used in individual studies must be con-sidered: more relevant velocity measurements can beobtained from TCCS with a corrected insonation angle.In extracerebral arteries, it is reasonable to assume thatchanges in velocity are proportional to changes in flow.This assumption cannot be extended to cerebral ves-sels due to their vasoreactivity, i.e., high velocity invasospasm with poor distal perfusion.
The velocity and indices of brain vessels are the moststudied parameter in AD and MCI patients. Unfor-tunately, the majority of published studies used theless-precise TCD system and there are no universallyapplicable norms for determining specific cognitivepathologies from velocity alone.
The most commonly reported finding is significantlylower flow velocities in AD patients compared withcontrols, particularly in middle cerebral artery (MCA),without a significant side asymmetry. The velocity thatwas most commonly decreased in MCA in AD patientscompared with healthy controls was the mean flowvelocity [6, 17–20]. This finding can be explained bythe fact that the MCA is the artery supplying the tempo-ral and parietal lobes most affected in AD. Accordingto a large longitudinal population study [21], sub-jects with higher MCA velocities were less likely todevelop AD. One available meta-analysis of 12 stud-ies found a lower CBF velocity and higher PI in ADand VaD patients compared with healthy, age-matchedcontrols. The majority of studies, however, reported nosignificant difference in flow velocities between indi-viduals with AD and VaD, respectively [22]. The othermajor cerebral arteries were studied less frequently andyielded ambiguous results [23–25].
Data from the opposite side of the age spectrum (pre-mature infants) support that lower TCD flow velocitiesare associated with poor cognitive performance [26].
In the context of this review, a comparison of healthysubjects and patients with MCI would be interestingbut the available results are ambiguous, likely due tothe insufficient sample size for detecting smaller dif-ferences than in AD [6, 24, 27].
CEREBROVASCULAR RESERVECAPACITY
CVR is a parameter of cerebrovascular autoregula-tion describing the ability for vasodilation of cerebralarterioles in setting of low cerebral perfusion pressure.The vasoactive stimuli used in testing CVR are inducedreductions in systemic blood pressure, the administra-tion of vasoactive substances such as acetazolamide,and changes in arterial CO2 levels. The main advan-tage claimed for acetazolamide is its independence ofsubject cooperation. However, there is considerableindividual variability in the response to acetazolamide.Both the serum acetazolamide concentrations for agiven dose, and the cerebrovascular responses to agiven serum level, vary between subjects [28]. Themost often-used experimental stimulus is the changein the arterial CO2 level. Hypercapnia can be inducedby direct CO2 inhalation or by breath-holding method,although in breath-holding there is also hypoxemiamodulating cerebrovascular responses to hypercapnia.CVR is then expressed as the ratio of the MV in basalconditions and the MV under the conditions of a higherCO2 level. In healthy brains, there is a physiologicalincrease of flow velocity. When breath-holding is thestimulus, the ratio can be multiplied by the duration ofbreath holding and expressed as a breath-holding index(BHI) [29]. Cut point of the BHI value for distinguish-ing between pathological and normal cerebrovascularreactivity in patients who have asymptomatic severeICA stenoses and symptomatic occlusions was deter-mined to be 0.69 [30]. In healthy volunteers, BHIvalues remain in the range between 1.03 and 1.65 [31].BHI is a linear index, therefore there is no differencebetween short (<27 s) and long (>27 s) measurementtimes [29]. The breath-holding procedure is prone toinaccuracy, particularly in older patients, and shouldbe performed with coregistration of end-tidal CO2(EtCO2) and blood pressure. Possible confounding dueto the Valsalva effect can be overcome by performingbreath holds after expiration. The measurement shouldbe repeated and averaged values should be adopted foreach hemisphere [32]. CVR decreases with increasingage [33] and reflects a generally impaired vascular sys-tem, leading not just to brain hypoperfusion but also toan increased overall mortality [34].
CVR results concerning cognitive decline are moreconsistent than those for flow velocities. The CVR ofMCA to hypercapnic [17, 18, 35, 36] and hypocap-nic [18, 37] stimuli in AD patients is significantlylower than in healthy controls. Only some studieswith insufficient statistical power do not fully support
AU
THO
R C
OP
Y
A. Tomek et al. / Utility of Transcranial Ultrasound S369
these findings [38, 39]. Cognitively intact subjects withhigher CVRs in MCA are less likely to develop AD orVaD [21]. MCI patients with pathological values ofthe BHI have greater risk of developing dementia thanpatients with normal BHI values [27]. In longitudi-nal follow-up studies, the BHI significantly correlatedwith neuropsychological tests: the Mini-Mental StateExamination and the Alzheimer’s Disease AssessmentScale-Cognitive subscale in AD [40]. Although theimpairment of CVR is more serious in VaD, it seemsthat the microvasculature is altered in both types ofdementia.
CVR can be sometimes confused with neurovascu-lar coupling (a process in which activation of a brainregion evokes a local increase in blood flow) althoughthese two phenomena are probably mediated by dif-ferent mechanisms [41]. Change in flow velocities inposterior cerebral artery in reaction to a visual stim-ulus was tested in an effort to differentiate betweenAD and VaD based on the fact that the occipital lobeshould be preserved until the late stages of AD com-pared with the early stages of VaD. The results of suchstudies were inconclusive and the observed differencewas insubstantial and depended on sample size [23,42–44]. Similar study in MCA region gave positiveresults when a reaction to a cognitive task was tested(side-specific activation in healthy controls, bilateraland lower increase in AD), but not in case of a reactionto a hand movement (no difference between AD andcontrols) [39].
The cerebral autoregulation in its classic definition(a mechanism that stabilizes CBF in the context ofchanges in blood pressure [45]) should be differenti-ated from CVR. It can be tested by various methodsincluding PET and TCD and, unlike CVR, it seems tobe preserved in AD patients [32, 46–48]. These find-ings suggest that in AD there is rather microvasculardysfunctions than impairment of larger vessels that areresponsible for cerebral autoregulation [49].
Chronic silent cerebral microembolization can causecognitive decline. Severe carotid stenosis or occlusion,which is thought to be linked to microembolization,can be associated with cognitive decline in otherwiseasymptomatic patients [50–52]. In this case, a role isnot only played by microembolization, but chronic
hypoperfusion as well [53, 54]. Carotid endarterec-tomy or stenting can improve but also worsen thecognitive decline [55]. The exacerbation is associ-ated with microembolization during the procedure andwith white-matter hyperintensities in MR diffusion-weighted images post-operatively [56]. The higherrate of cerebral microembolization and consequentlycognitive decline was found also during coronary inter-ventions [57].
The presence of spontaneous cerebral microembolican be assessed using long-term TCD monitoring. Ahead frame with ultrasound probes is attached to thehead of the patient and the continuity of flow in bothMCA arteries is monitored over the course of typicallyone hour. The microembolization is recognized as ahigh-intensity transient signal (HITS). By counting thefrequency of HITSs, the severity of microembolizationcan be estimated.
There have been many studies related to cerebralmicroembolization in carotid stenosis, during carotidor coronary interventions, and during valve replace-ments, but only a limited number of works haveassessed the presence or severity of microembolizationin AD. According to this limited evidence, the inci-dence of microembolization is higher in patients withdementia (both in AD and VaD) than in healthy con-trols [58, 59]. In patients with microembolization, theprogression of cognitive decline was faster in 6-monthintervals over the course of 1.5 years and the depres-sive symptoms were more prominent than in patientswithout microembolization [60].
Spontaneous cerebral microembolization can occurnot only in the settings of arterial disease or diseases ofthe left heart (artificial cardiac valves, hypokinesis ofcardiac walls after myocardial infarction, etc.), but alsoin the settings of venous disease in combination withright-left shunts (atrial septal defects, foramen ovalepatens, or extracardiac pulmonary shunts). Right-leftshunts can be assessed by ultrasound monitoring ofHITSs after intravenous injection of a microbubblesuspension (hydroxyethylstarch, agitated saline). Theassessment is performed during rest and during the Val-salva maneuver that reverses the pressure in the rightand left parts of the heart and enables more bubbles toescape through the septal defect from the right atriumto the left atrium. The severity of the shunt is estimatedbased on the amount of HITSs. The accuracy of thismethod is comparable to or even higher than that of thetransesophageal echocardiography, which is used as agold standard [61].
In the first pilot study [58], there was a signifi-cant difference in the incidence of right-left shunts
AU
THO
R C
OP
Y
S370 A. Tomek et al. / Utility of Transcranial UltrasoundTa
ble
1O
verv
iew
ofT
CD
stud
ies
inde
men
tiapa
tient
s
Aut
hor
Met
hods
Popu
latio
nPa
ram
eter
sst
udie
dR
esul
ts
Asi
l[44
]C
ompa
riso
nof
reac
tivity
tovi
sual
stim
ulus
inoc
cipi
tall
obes
inA
D,V
aD,h
ealth
yco
ntro
ls(P
CA
bila
t.,ey
escl
osed
/eye
sop
ened
)
n=
36(1
5A
D,1
2V
aD,9
cont
rols
)M
FV(P
CA
),pe
rcen
tage
velo
city
vari
atio
nsN
osi
gnifi
cant
diff
eren
cein
flow
velo
citie
sat
rest
orst
imul
ibet
wee
ngr
oups
;dec
reas
edre
activ
ityin
VaD
Bar
[35]
Com
pari
son
ofC
VR
inA
D,V
aDan
dhe
alth
yco
ntro
ls,a
sses
smen
tof
the
effe
ctof
AC
hEI
trea
tmen
t(M
CA
bila
t.,ba
salc
ondi
tions
/5%
CO
2
inha
latio
n),r
epea
ted
afte
r5
wee
ksof
gala
ntam
ine
n=
54(1
7A
D,1
7V
aD,2
0co
ntro
ls)
MFV
(MC
A),
perc
enta
geve
loci
tyva
riat
ions
CV
Rin
MC
Ade
crea
sed
inA
Dan
dV
aDin
com
pari
son
tohe
alth
yco
ntro
ls,b
ette
rC
VR
afte
rga
lant
amin
etr
eatm
ento
nbo
thA
Dan
dV
aD
Caa
man
o[2
5]C
ompa
riso
nof
flow
velo
citie
sin
AD
,VaD
and
cont
rols
(MC
A,B
A,b
asal
cond
ition
s)n
=36
(12
AD
,12
VaD
,12
cont
rols
)M
FV(M
CA
,BA
)D
ecre
ased
valu
esin
dem
ente
dpa
tient
s
Cla
asse
n[1
9]C
ompa
riso
nof
flow
velo
citie
sin
AD
and
heal
thy
cont
rols
(MC
Aon
esi
de,b
asal
cond
ition
s)n
=17
(9A
D,8
cont
rols
)M
FV(M
CA
)Si
gnifi
cant
lyre
duce
dflo
wve
loci
tyin
AD
Gho
rban
i[66
]C
ompa
riso
nof
flow
velo
citie
sin
AD
trea
ted
and
non-
trea
ted
bydo
nepe
zil(
MC
Aan
dPC
Abi
lat.,
repe
ated
afte
r4
wee
ksof
done
pezi
l5m
gan
dan
othe
r4
wee
ksof
done
pezi
l10
mg)
n=
22(1
1A
D,1
1co
ntro
ls)
PSV
,MFV
,ED
V(P
CA
,M
CA
)In
crea
sein
PSV
and
MFV
inM
CA
,and
MFV
and
ED
Vin
PCA
afte
r10
mg
trea
tmen
t
Guc
uyen
er[2
3]C
ompa
riso
nof
flow
velo
citie
san
dre
activ
ityto
visu
alst
imul
usin
occi
pita
llob
esin
AD
and
depr
essi
veps
eudo
dem
entia
(PC
Abi
lat.,
basa
lco
nditi
ons/
visu
alst
imul
us)
n=
34(1
1A
D,1
3V
aD,1
0co
ntro
ls)
MFV
(PC
A),
perc
enta
geve
loci
tyva
riat
ions
Low
erflo
wve
loci
ties
atre
stan
daf
ter
stim
ulus
inbo
thA
Dan
dde
pres
sive
pseu
dode
men
tiath
enco
ntro
ls.R
eact
ivity
tovi
sual
stim
ulus
impa
ired
inA
D,n
otin
depr
essi
veps
eudo
dem
entia
Lee
[36]
Ass
essm
ento
fflo
wve
loci
ties
and
CV
Rin
AD
(MC
Abi
lat.,
basa
lcon
ditio
ns/5
min
ofre
brea
thin
g)
n=
34(1
7A
D,1
7co
ntro
ls)
MFV
,PI
(MC
A),
perc
enta
geve
loci
tyva
riat
ions
No
diff
eren
cein
base
line
MFV
and
PIbe
twee
nsu
bjec
tsan
dco
ntro
ls,C
VR
sign
ifica
ntly
decr
ease
don
both
side
sin
AD
Lik
itjar
oen
[38]
Com
pari
son
ofC
VR
inA
Dan
dV
aD(M
CA
bila
t.,ba
salc
ondi
tions
/2,5
,10,
20m
inaf
ter
1,00
0m
gac
etaz
olam
ide
i.v.)
n=
18(9
AD
,9V
aD)
PSV
,MFV
,ED
V(M
CA
),pe
rcen
tage
velo
city
vari
atio
ns
Not
sign
ifica
ntre
sults
(hig
her
ED
Van
dM
FV,l
ower
PSV
atba
salc
ondi
tions
inA
Dth
anV
aD,b
ette
rC
VR
inA
Dth
anV
aD)
Mat
teis
[39]
Com
pari
son
ofC
VR
and
reac
tivity
tom
otor
and
cogn
itive
task
inA
Dan
dV
aD(M
CA
bila
t.,ba
sal
cond
ition
s/ap
nea/
hand
mov
emen
t/ver
bala
ndde
sign
disc
rim
inat
ion)
n=
40(1
0A
D,1
0V
aD,1
0co
ntro
ls)
MFV
(MC
A),
perc
enta
geve
loci
tyva
riat
ions
CV
Rto
apne
alo
wer
inV
aD;h
and
mov
emen
t–co
ntra
late
rali
ncre
ase
inflo
wve
loci
ties
inA
Dan
dco
ntro
ls,b
ilate
rali
nV
aD;b
ilate
ralr
espo
nse
onco
gniti
vest
imul
iin
AD
and
VaD
,cor
resp
ondi
ngsi
dere
spon
sein
cont
rols
Prov
inci
ali[
37]
Com
pari
son
ofC
VR
inA
D,V
aDan
dco
ntro
ls(M
CA
,bas
alco
nditi
ons/
hype
rven
tilat
ion/
long
est
poss
ible
apne
a/5
min
air
rebr
eath
ing)
n=
65(2
0A
D,2
0V
aD,2
5co
ntro
ls)
MFV
,PI
(MC
A),
perc
enta
geve
loci
tyva
riat
ions
Hig
her
PI,l
ower
velo
city
decr
ease
inhy
perv
entil
atio
nin
both
dem
entia
s;re
stflo
wve
loci
ties
and
resp
onse
tohy
perc
apni
alo
wer
inV
aDth
anA
Dor
cont
rols
Pura
ndar
e[5
8]C
ompa
riso
nof
spon
tane
ous
cere
bral
mic
roem
boli,
veno
us-t
o-ar
teri
alci
rcul
atio
nsh
unts
and
caro
tidar
tery
dise
ase
inA
D,V
aDan
dco
ntro
ls(M
CA
bila
t.)
57(2
4A
D,1
7V
aD,
16co
ntro
ls)
Spon
tane
ous
cere
bral
mic
roem
boli
(MC
A),
bubb
les
(MC
A)
Mor
ece
rebr
alm
icro
embo
liin
VaD
than
cont
rols
,in
AD
nots
igni
fican
t,no
diff
eren
cein
veno
us-t
o-ar
teri
alci
rcul
atio
nsh
unts
orca
rotid
sten
osis
betw
een
dem
entia
and
cont
rols
Pura
ndar
e[5
9]Sp
onta
neou
sce
rebr
alm
icro
embo
li,ri
ght-
left
circ
ulat
ion
shun
tsan
dca
rotid
arte
rydi
seas
ein
AD
,VaD
and
cont
rols
(MC
Abi
lat.)
320
(85
AD
,85
VaD
,15
0co
ntro
ls)
Spon
tane
ous
cere
bral
mic
roem
boli
(MC
A),
bubb
les
(MC
A)
Mor
ece
rebr
alm
icro
embo
liin
VaD
and
AD
than
cont
rols
,no
diff
eren
cein
shun
tor
caro
tidst
enos
isbe
twee
nde
men
tiaan
dco
ntro
ls
AU
THO
R C
OP
Y
A. Tomek et al. / Utility of Transcranial Ultrasound S371
Tabl
e1
(Con
tinu
ed)
Aut
hor
Met
hods
Popu
latio
nPa
ram
eter
sst
udie
dR
esul
ts
Rie
s[2
2]C
ompa
riso
nof
flow
velo
citie
san
def
fect
ive
puls
atili
tyra
tein
AD
,VaD
and
heal
thy
cont
rols
(MC
A,A
CA
,PC
A,V
Abi
lat.,
BA
)
n=
105
(24
AD
,17
VaD
,64
cont
rols
)PS
V,M
FV,E
DV
(MC
A,
AC
A,P
CA
,VA
,BA
),ef
fect
ive
puls
atili
tyra
te
No
diff
eren
cein
PSV
inal
l3gr
oups
,sig
nific
antly
low
erin
MFV
,ED
Van
dhi
gher
effe
ctiv
epu
lsat
ility
rate
inV
aDco
mpa
red
toA
Dor
cont
rols
Roh
er[2
0]C
ompa
riso
nof
mea
nflo
wve
loci
ties
and
PIin
intr
acra
nial
arte
ries
inA
Dan
dhe
alth
yco
ntro
ls(1
6se
gmen
tsof
circ
leof
Will
is)
n=
55(2
5A
D,3
0co
ntro
ls)
MFV
(16
segm
ents
ofci
rcle
ofW
illis
)H
ighe
rPI
inA
D,n
on-s
igni
fican
t:ge
nera
llylo
wer
MFV
inA
D
Roh
er[6
]C
ompa
riso
nof
flow
velo
citie
san
dPI
inin
trac
rani
alar
teri
esin
AD
and
heal
thy
cont
rols
(16
segm
ents
ofci
rcle
ofW
illis
)
n=
103
(42
AD
,11
VaD
,50
cont
rols
)PS
V,M
FV,E
DV
(16
segm
ents
ofci
rcle
ofW
illis
)
Sign
ifica
ntly
low
erM
FVan
dhi
gher
PIin
left
siph
on,l
eftI
CA
and
righ
tdis
talM
CA
inA
Dth
anin
heal
thy
cont
rols
Ros
enga
rten
[42]
Com
pari
son
offlo
wve
loci
ties
and
reac
tivity
tovi
sual
stim
ulus
inA
Dan
dhe
alth
yco
ntro
ls,
influ
ence
ofA
ChE
Itr
eatm
enti
nA
D(l
eftP
CA
,ri
ghtM
CA
,bas
alco
nditi
ons/
text
read
ing)
,re
peat
edaf
ter
4w
eeks
ofdo
nepe
zil5
mg
and
anot
her
4w
eeks
ofdo
nepe
zil1
0m
g
n=
24(8
AD
,16
VaD
)PS
V(P
CA
,MC
A),
perc
enta
geve
loci
tyva
riat
ions
Low
erve
loci
ties
and
incr
ease
dst
iffn
ess
inA
Dth
anhe
alth
yco
ntro
ls,d
ose
depe
nden
tnor
mal
izat
ion
ofbo
thpa
ram
eter
saf
ter
done
pezi
l
Ros
enga
rten
[43]
Com
pari
son
offlo
wve
loci
ties
and
reac
tivity
tovi
sual
stim
ulus
inoc
cipi
tall
obes
inA
D,V
aDan
dco
ntro
ls(P
CA
,eye
scl
osed
/stim
ulat
ion
byvi
sual
stim
ulus
)
n=
40(1
5A
D,1
0V
aD,1
5co
ntro
ls)
PSV
(PC
A),
perc
enta
geve
loci
tyva
riat
ions
No
sign
ifica
ntdi
ffer
ence
sin
rest
ing
flow
velo
citie
s,lo
wer
incr
ease
inPS
Vin
VaD
com
pare
dto
AD
and
cont
rols
,non
-sig
nific
antr
esul
tsin
AD
Rui
tenb
erg
[21]
Cor
rela
tion
offlo
wve
loci
ties
with
mar
kers
ofco
gniti
vede
clin
ean
dhi
ppoc
ampa
latr
ophy
(MC
Abi
lat.,
rest
/5m
inof
5%C
O2)
n=
1,73
2(1
3A
D,1
VaD
,171
8he
alth
y)PS
V,M
V,E
DV
(MC
A)
Gre
ater
PSV
,MFV
,ED
V–
less
likel
yde
men
tiaan
dbi
gger
hipp
ocam
pus
and
amyg
dala
;No
asso
ciat
ion
ofC
VR
and
pres
ence
ofde
men
tiaSi
lves
trin
i[40
]C
orre
latio
nof
chan
ges
inflo
wve
loci
ties
and
BH
Iw
ithpr
ogre
ssio
nof
cogn
itive
impa
irm
ent(
MC
Abi
lat.,
basa
lcon
ditio
ns/b
reat
h-ho
ldin
g),r
epea
ted
afte
r12
mon
thof
done
pezi
l(5
mg
daily
for
3m
onth
,the
n10
mg
daily
)
n=
53(5
3A
D)
MFV
(MC
Abi
lat.)
Posi
tive
corr
elat
ion
ofne
urop
sych
olog
ical
test
sch
ange
sw
ithB
HI,
age
and
DM
Stef
ani[
18]
Com
pari
son
ofce
rebr
alhe
mod
ynam
ics
inA
Dan
dco
ntro
ls(M
CA
bila
t.,ba
sal
cond
ition
s/br
eath
-hol
ding
)
n=
80(4
0A
D,4
0co
ntro
ls)
MFV
,PI,
BH
I(M
CA
)L
ower
MFV
,hig
her
PIan
dlo
wer
BH
Iin
AD
than
inco
ntro
ls
Sun
[24]
Com
pari
son
offlo
wve
loci
ties
inM
CI
and
cont
rols
,co
mpa
riso
nof
flow
velo
citie
sin
MC
IA
poE
ε4
carr
iers
and
non-
carr
iers
(AC
A,M
CA
bila
t.,B
A)
n=
60(3
0M
CI,
30co
ntro
ls)
PSV
,MFV
,ED
V(A
CA
,M
CA
,BA
)L
ower
PSV
,MFV
and
ED
Vin
MC
Aan
dA
CA
bila
t.in
MC
Ith
anin
cont
rols
;low
erPS
V,M
FV,E
DV
inM
CA
bila
t.(e
xcep
tof
PSV
inle
ftM
CA
)in
MC
IA
poE
ε4
carr
iers
than
non-
carr
iers
Vic
enzi
ni[1
7]C
ompa
riso
nof
flow
velo
citie
s,PI
and
CV
Rin
AD
,V
aD,a
ndco
ntro
ls(M
CA
bila
t.,ba
sal
cond
ition
s/hy
perv
entil
atio
n/C
O2
inha
latio
n)
n=
180
(60
AD
,58
VaD
,62
cont
rols
)PS
V,M
FV,E
DV
,PI
(MC
A)
Low
erM
FV,h
ighe
rPI
and
low
erC
VR
inA
Dan
dV
aDco
mpa
red
toco
ntro
ls
Viti
cchi
[27]
Com
pari
son
ofIM
Tan
dB
HI
inA
Dan
dM
CI,
asso
ciat
ion
ofth
ese
para
met
ers
with
the
risk
ofco
nver
sion
from
MC
Ito
AD
(IM
Tin
caro
tids,
BH
Iin
MC
Abi
lat.)
n=
117
(117
MC
I)B
HI
(MC
A)
Hig
her
IMT
and
low
erB
HI
inA
Dth
anM
CI,
asso
ciat
ion
ofhi
gher
IMT
and
low
erB
HI
with
fast
erpr
ogre
ssio
nfr
omM
CI
tode
men
tia
AC
A,
ante
rior
cere
bral
arte
ry;
AC
hEI,
acet
ylch
olin
ees
tera
sein
hibi
tor;
AD
,A
lzhe
imer
’sdi
seas
e;B
A,
basi
lar
arte
ry;
BH
I,br
eath
-hol
ding
inde
x;C
VR
,ce
rebr
ovas
cula
rre
serv
eca
paci
ty;
DM
,di
abet
esm
ellit
us;E
DV
,end
dias
tolic
velo
city
;IC
A,i
nter
nalc
arot
idar
tery
;IM
T,in
tima-
med
iath
ickn
ess;
MC
A,m
iddl
ece
rebr
alar
tery
;MC
I,m
ildco
gniti
veim
pair
men
t;M
FV,m
ean
flow
velo
city
;M
V,m
ean
flow
velo
city
;PC
A,p
oste
rior
cere
bral
arte
ry;P
I,pu
lsat
ility
inde
x;PS
V,p
eak
syst
olic
velo
city
;VA
,ver
tebr
alar
tery
;VaD
,vas
cula
rde
men
tia.
AU
THO
R C
OP
Y
S372 A. Tomek et al. / Utility of Transcranial Ultrasound
between individuals with dementia (both AD and VaD)and healthy controls. However, the results were notconfirmed by the later case-control study [59]. Theprevalence of intracardiac right-left shunts in the studygroup was 29% in individuals with AD and 32% in indi-viduals with VaD, which is slightly higher than usuallyreported; 20–25% in the general population [62].
ADVANTAGES AND DISADVANTAGES OFTRANSCRANIAL ULTRASOUND
Sufficient evidence exists to support the screeningrole of transcranial ultrasound in cognitive medicine.Notwithstanding, it is important to emphasize that theeffectiveness of vascular factor control strategies inslowing cognitive decline is not supported by sufficientevidence in the form of randomized clinical trials.
Transcranial ultrasound provides an advantage overother techniques of assessing vascular pathology, espe-cially elsewhere in the body (extremities, heart),because it directly informs about the vessels in thetarget organ—the brain. TCD examination can leadto incident discovery and specific preventive measuresof intracranial atherosclerosis, associated with a par-ticularly high risk of stroke [63]. The fact that TCDis conducted by a cerebrovascular expert can lead toa better standard of care in patients with concomi-tant cerebrovascular and AD, leading possibly to anincreased reduction in vascular morbidity and mor-tality. Even if the vascular hypothesis of AD is notvalid, vascular preventive interventions can lead to areduction in overall vascular morbidity and mortality.
The main disadvantage of TCD arises from its insuf-ficient evidence base; randomized clinical trials areurgently needed to prove its utility. The dependenceon an operator experience and a good patient cooper-ation especially when performing the breath holdingmaneuver must be mentioned. A promising avenueof research in the near future is the combinationof TCD with other techniques such as near-infraredspectroscopy [64, 65]. TCD cannot currently help indifferentiating AD and VaD; this research aim shouldbe investigated with further work.
CONCLUSIONS
On the basis of current knowledge, we recommenda standard ultrasound examination of extracranialand intracranial cerebral vessels for signs of possiblestenotic or occlusive cerebrovascular disease in MCIand AD patients. This examination could be comple-
mented with a dedicated screening test—assessmentof CVR with BHI.
Surveillance and treatment of vascular risk factors inthe preclinical stages of AD are of significant clinicalimportance and could help to delay the development ofcognitive decline in susceptible individuals. Because aclinically approved cure for AD seems unlikely in thenear future, preventions aimed specifically at individ-uals with vascular risk factors appear to be one of themost reasonable alternatives.
ACKNOWLEDGMENTS
The authors of the article were supported by theGrant Agency of Ministry of Health of Czech Republic(IGA No. NT/13319).
DISCLOSURE STATEMENT
Authors’ disclosures available online (http://www.j-alz.com/disclosures/view.php?id=2578).
REFERENCES
[1] de la Torre J, Mussivand T (1993) Can disturbed brainmicrocirculation cause Alzheimer’s disease? Neurol Res 15,146-153.
[2] Weller RO, Subash M, Preston SD, Mazanti I, Carare RO(2007) Clearance of A� from the brain in Alzheimer’s disease:Perivascular drainage of amyloid-� peptides from the brainand its failure in cerebral amyloid angiopathy and Alzheimer’sdisease. Brain Pathol 18, 253-266.
[3] Pimentel-Coelho PM, Rivest S (2012) The early contributionof cerebrovascular factors to the pathogenesis of Alzheimer’sdisease. Eur J Neurosci 35, 1917-1937.
[4] Kokjohn TA, Maarouf CL, Roher AE (2012) Is Alzheimer’sdisease amyloidosis the result of a repair mechanism goneastray? Alzheimers Dement 8, 574-583.
[5] van Es A, van der Grond J, ten Dam V, de Craen A, BlauwG, Westendorp R (2010) Associations between total cerebralblood flow and age related changes of the brain. PLoS One 5,e9825.
[6] Roher A, Garami Z, Tyas S, Maarouf C, Kokjohn T,Belohlavek M, Vedders L, Connor D, Sabbagh M, Beach T,Emmerling M (2011) Transcranial doppler ultrasound bloodflow velocity and pulsatility index as systemic indicators forAlzheimer’s disease. Alzheimers Dement 7, 445-455.
[7] Dichgans M, Zietemann V (2012) Prevention of vascular cog-nitive impairment. Stroke 43, 3137-3146.
[11] Vlaar AMM, Bouwmans A, Mess WH, Tromp SC, WeberWEJ (2009) Transcranial duplex in the differential diagnosisof parkinsonian syndromes. J Neurol 256, 530-538.
[12] Michel E, Zernikow B (1998) Gosling’s Doppler pulsatilityindex revisited. Ultrasound Med Biol 24, 597-599.
[13] de Riva N, Pudohoski K, Smielewski P, Kasprowicz M,Zweifel C, Steiner L, Reinhard M, Fabregas N, PickardJ, Czosnyka M (2012) Transcranial doppler pulsatilityindex: What it is and what it isn’t. Neurocrit Care 17,58-66.
[14] Bude R, Rubin J (1999) Relationship between the resistiveindex and vascular compliance and resistance. Radiology 211,411-417.
[15] Bragoni M, Feldman E (1996) Transcranial Doppler indicesof intracranial hemodynamics. In Neurosonology, TegelerC, Babikian V, Gomez C, eds. Mosby, St. Louis, pp. 129-139.
[16] Krejza J, Mariak Z, Walecki J, Szydlik P, Lewko J, Ustymow-icz A (1999) Transcranial color Doppler sonography of basalcerebral arteries in 182 healthy subjects: Age and sex variabil-ity and normal reference values for blood flow parameters. AmJ Roentgenol 172, 213-218.
[17] Vicenzini E, Ricciardi M, Altieri M, Puccinelli F, BonaffiniN, Di Piero V, Lenzi G (2007) Cerebrovascular reactivity indegenerative and vascular dementia: A transcranial Dopplerstudy. Eur Neurol 58, 84-89.
[18] Stefani A, Sancesario G, Pierantozzi M, Leone G, Galati S,Hainsworth A, Diomedi M (2009) CSF biomarkers, impair-ment of cerebral hemodynamics and degree of cognitivedecline in Alzheimer’s and mixed dementia. J Neurol Sci 283,109-115.
[19] Claassen J, Diaz-Arrastia R, Martin-Cook K, Levine B, ZhangR (2009) Altered cerebral hemodynamics in early Alzheimerdisease: A pilot study using transcranial Doppler. J AlzheimersDis 17, 621-629.
[20] Roher A, Garami Z, Alexandrov A, Kokjohn T, Esh C,Kalback W, Vedders L, Wilson J, Sabbagh M, Beach T (2006)Interaction of cardiovascular disease and neurodegeneration:Transcranial Doppler ultrasonography and Alzheimer’s dis-ease. Neurol Res 28, 672-678.
[21] Ruitenberg A, den Heijer T, Bakker S, van Swieten J, Koud-staal P, Hofman A, Breteler M (2005) Cerebral hypoperfusionand clinical onset of dementia: The Rotterdam Study. AnnNeurol 57, 789-794.
[22] Ries F, Horn R, Hillekamp J, Honisch C, Konig M, Soly-mosi L (1993) Differentiation of multi-infarct and Alzheimerdementia by intracranial hemodynamic parameters. Stroke 24,228-235.
[23] Gucuyener D, Yenilmez C, Ayranci U, Ozdemir F, UzunerN, Ozkan S, Kaptanoglu C, Ozdemir G (2010) An analysisof changes in cerebral blood flood velocities in depressivepseudo-dementia and Alzheimer disease patients. Neurologist16, 358-363.
[24] Sun ZW, Zhu YX, Liu HY, Liu J, Zhu XQ, Zhou JN, Liu RY(2007) Decreased cerebral blood flow velocity in apolipopro-tein E ?4 allele carriers with mild cognitive impairment. EurJ Neurol 14, 150-155.
[25] Caamano J, Gomez M, Cacabelos R (1993) TranscranialDoppler ultrasonography in senile dementia: Neuropsycho-logical correlations. Methods Find Exp Clin Pharmacol 15,193-199.
[26] Bakker MJ, Hofmann J, Churches OF, Badcock NA, KohlerM, Keage HA (2014) Cerebrovascular function and cogni-
tion in childhood: A systematic review of transcranial dopplerstudies. BMC Neurol 14, 1-12.
[27] Viticchi G, Falsetti L, Vernieri F, Altamura C, Bartolini M,Luzzi S, Provinciali L, Silvestrini M (2012) Vascular pre-dictors of cognitive decline in patients with mild cognitiveimpairment. Neurobiol Aging 33, e1121-e1129.
[28] Fierstra J, Sobczyk O, Battisti-Charbonney A, Mandell DM,Poublanc J, Crawley AP, Mikulis DJ, Duffin J, Fisher JA(2013) Measuring cerebrovascular reactivity: What stimulusto use? J Physiol 591, 5809-5821.
[29] Markus HS, Harrison MJ (1992) Estimation of cerebrovascu-lar reactivity using transcranial Doppler, including the use ofbreath-holding as the vasodilatory stimulus. Stroke 23, 668-673.
[31] Zavoreo I, Basic Kes V, Coric L, Demarin V (2012) Breathholding index and arterial stiffness in evaluation of stroke riskin diabetic patients. Perspect Med 1, 156-159.
[32] van Beek AH, de Wit HM, Olde Rikkert MG, Claassen JA(2011) Incorrect performance of the breath hold method inthe old underestimates cerebrovascular reactivity and goesunnoticed without concomitant blood pressure and end-tidalCO2 registration. J Neuroimag 21, 340-347.
[33] Peisker T, Bartos A, Skoda O, Ibrahim I, Kalvach P (2010)Impact of aging on cerebral vasoregulation and parenchymalintegrity. J Neurol Sci 299, 112-115.
[34] Portegies ML, de Bruijn RF, Hofman A, Koudstaal PJ, IkramMA (2014) Cerebral Vasomotor Reactivity and Risk of Mor-tality The Rotterdam Study. Stroke 45, 42-47.
[35] Bar K, Boettger M, Seidler N, Mentzel H, Terborg C, SauerH (2007) Influence of galantamine on vasomotor reactivityin Alzheimer’s disease and vascular dementia due to cerebralmicroangiopathy. Stroke 38, 3186-3192.
[36] Lee S-T, Jung K-H, Lee Y-S (2007) Decreased vasomo-tor reactivity in Alzheimer’s disease. J Clin Neurol 3,18-23.
[37] Provinciali L, Minciotti P, Ceravolo G, Angeleri F, SanguinettiC (1990) Transcranial Doppler sonography as a diagnostictool in vascular dementia. Eur Neurol 30, 98-103.
[38] Likitjaroen Y, Suwanwela N, Phanthumchinda K (2009)Vasoreactivity induced by acetazolamide in patients with vas-cular dementia versus Alzheimer’s disease. J Neurol Sci 283,32-35.
[39] Matteis M, Silvestrini M, Troisi E, Bragoni M, Vernieri F,Caltagirone C (1998) Cerebral hemodynamic patterns duringstimuli tasks in multi-infarct and Alzheimer types of demen-tia. Acta Neurol Scand 97, 374-380.
[40] Silvestrini M, Pasqualetti P, Baruffaldi R, Bartolini M, Han-douk Y, Matteis M, Moffa F, Provinciali L, Vernieri F (2006)Cerebrovascular reactivity and cognitive decline in patientswith Alzheimer disease. Stroke 37, 1010-1015.
[41] Lin A, Fox P, Hardies J, Duong T, Gao J (2010) Nonlinearcoupling between cerebral blood flow, oxygen consumption,and ATP production in human visual cortex. Proc Natl AcadSci U S A 107, 8446-8451.
[43] Rosengarten B, Paulsen S, Molnar S, Kaschel R, GallhoferB, Kaps M (2007) Activation-flow coupling differentiatesbetween vascular and Alzheimer type of dementia. J NeurolSci 257, 149-154.
AU
THO
R C
OP
Y
S374 A. Tomek et al. / Utility of Transcranial Ultrasound
[44] Asil T, Uzuner N (2005) Differentiation of vasculardementia and Alzheimer disease: A functional transcranialDoppler ultrasonographic study. J Ultrasound Med 24, 1065-1070.
[45] van Beek A, Claassen J, Rikkert M, Jansen R (2008) Cerebralautoregulation: An overview of current concepts and method-ology with special focus on the elderly. J Cereb Blood FlowMetab 28, 1071-1085.
[46] Zazulia A, Videen T, Morris J, Powers W (2010) Autoregu-lation of cerebral blood flow to changes in arterial pressurein mild Alzheimer’s disease. J Cereb Blood Flow Metab 30,1883-1889.
[47] van Beek A, Sijbesma J, Jansen R, Rikkert M, Claassen J(2010) Cortical oxygen supply during postural hypotensionis further decreased in Alzheimer’s disease, but unrelated tocholinesteraseinhibitor use. J Alzheimers Dis 21, 519-526.
[48] Gommer E, Martens E, Aalten P, Shijaku E, Verhey F, Mess W,Ramakers I, Reulen J (2012) Dynamic cerebral autoregulationin subjects with Alzheimer’s disease, mild cognitive impair-ment, and controls: Evidence for increased peripheral vascularresistance with possible predictive value. J Alzheimers Dis 30,805-813.
[50] Johnston S, O’Meara E, Manolio T, Lefkowitz D, O’LearyD, Goldstein S, Carlson M, Fried L, Longstreth WJ (2004)Cognitive impairment and decline are associated with carotidartery disease in patients without clinically evident cere-brovascular disease. Ann Intern Med 140, 237-247.
[51] Chang X, Zhou H, Lei C, Wu B, Chen Y, Hao Z, Dong W, LiuM (2013) Association between asymptomatic carotid steno-sis and cognitive function: A systematic review. NeurosciBiobehav Rev 37, 1493-1499.
[52] Balucani C, Viticchi G, Falsetti L, Silvestrini M (2012) Cere-bral hemodynamics and cognitive performance in bilateralasymptomatic carotid stenosis. Neurology 79, 1788-1795.
[53] Demarin V, Zavoreo I, Kes V (2012) Carotid artery diseaseand cognitive impairment. J Neurol Sci 322, 107-111.
[54] Sztriha L, Nemeth D, Sefcsik T, Vecsei L (2009) Carotidstenosis and the cognitive function. J Neurol Sci 283, 36-40.
[55] Gaudet J, Meyers P, McKinsey J, Lavine S, Gray W, MitchellE, Connolly EJ, Heyer E (2009) Incidence of moderate tosevere cognitive dysfunction in patients treated with carotidartery stenting. Neurosurgery 65, 325-329.
[56] Maggio P, Altamura C, Landi D, Migliore S, Lupoi D, MoffaF, Quintiliani L, Vollaro S, Palazzo P, Altavilla R, PasqualettiP, Errante Y, Quattrocchi C, Tibuzzi F, Passarelli F, ArpesaniR, di Giambattista G, Grasso F, Luppi G, Vernieri F (2013)Diffusion-weighted lesions after carotid artery stenting areassociated with cognitive impairment. J Neurol Sci 328, 58-63.
[57] Pugsley W, Klinger L, Paschalis C, Treasure T, HarrisonM, Newman S (1994) The impact of microemboli duringcardiopulmonary bypass on neuropsychological functioning.Stroke 25, 1393-1399.
[58] Purandare N, Welsh S, Hutchinson S, Riding G, Burns A,McCollum C (2005) Cerebral emboli and paradoxical emboli-sation in dementia: A pilot study. Int J Geriatr Psychiatry 20,12-16.
[59] Purandare N, Burns A, Daly KJ, Hardicre J, Morris J, Mac-farlane G, McCollum C (2006) Cerebral emboli as a potentialcause of Alzheimer’s disease and vascular dementia: Case-control study. BMJ 332, 1119-1124.
[60] Purandare N, Burns A, Morris J, Perry E, Wren J, McCol-lum C (2012) Association of cerebral emboli with acceleratedcognitive deterioration in Alzheimer’s disease and vasculardementia. Am J Psychiatry 169, 300-308.
[61] Nemec J, Marwick T, Lorig R, Davison M, Chimowitz M,Litowitz H, Salcedo E (1991) Comparison of transcranialDoppler ultrasound and transesophageal contrast echocardio-graphy in the detection of interatrial right-to-left shunts. AmJ Cardiol 68, 498-502.
[62] Hara H, Virmani R, Ladich E, Mackey-Bojack S, Titus J, Reis-man M, Gray W, Nakamura M, Mooney M, Poulose A (2005)Patent foramen ovale: Current pathology, pathophysiology,and clinical status. J Am Coll Cardiol 46, 1768-1776.
[63] Gorelick P, Wong K, Bae H, Pandey D (2008) Large arteryintracranial occlusive disease: A large worldwide burden buta relatively neglected frontier. Stroke 39, 2396-2399.
[64] Obrig H (2014) NIRS in clinical neurology—a promisingtool? Neuroimage 85, 535-546.
[65] Tarumi T, Dunsky DI, Khan MA, Liu J, Hill C, Arm-strong K, Martin-Cook K, Cullum CM, Zhang R (2014)Dynamic cerebral autoregulation and tissue oxygenation inamnestic mild cognitive impairment. J Alzheimers Dis 41,765-778.
[66] Ghorbani A, Chitsaz A, Shisbegar M, Akbari M (2010) Evalu-ation of the effect of donepezil on cerebral blood flow velocityin Alzheimer’s disease. Neurosciences 15, 172-176.
