Eur Radiol

DOI 10.1007/s00330-015-3957-z

NEURO

The impact of MRI combined with visual rating scales

on the clinical diagnosis of dementia: a prospective study
Martijn V. Verhagen 1 & Gerard L. Guit 1 & Gerrit Jan Hafkamp 2 &
Kees Kalisvaart 2

Received: 21 May 2015 / Revised: 20 July 2015 / Accepted: 29 July 2015

# European Society of Radiology 2015

Abstract
Objectives Dementia is foremost a clinical diagnosis.
However, in diagnosing dementia, it is advocated to perform
at least one neuroimaging study. This has two purposes: to
rule out potential reversible dementia (PRD), and to help determine the dementia subtype. Our first goal was to establish if
MRI combined with visual rating scales changes the clinical
diagnosis. The second goal was to demonstrate if MRI contributes to a geriatricians confidence in the diagnosis.
Methods The dementia subtype was determined prior to and
after MRI. Scoring scales used were: global cortical atrophy
(GCA), medial temporal atrophy (MTA), and white matter
hyperintensity measured according to the Fazekas scale. The
confidence level of the geriatrician was determined using a
visual analogue scale.
Results One hundred and thirty-five patients were included. After MRI, the diagnosis changed in 23.7 % (CI
17.0 %-31.1 %) of patients. Change was due to vascular
aetiology in 13.3 % of patients. PRD was found in 2.2 % of
all patients. The confidence level in the diagnosis increased
significantly after MRI (p = 0.001).
Conclusions MRI, combined with visual rating scales, has a
significant impact on dementia subtype diagnosis and on a
geriatricians confidence in the final diagnosis.

Key points
MRI with visual rating scales changes the dementia subtype
diagnosis significantly.
MRI is essential in demonstrating vascular disease as a
cause of dementia.
All suspected dementia patients should undergo an MRI with
visual rating scales.
MRI improves a geriatricians confidence in the diagnosis of
the dementia subtype.
MRI remains essential during the workup of dementia to
exclude reversible causes.
Keywords Dementia . Magnetic resonance imaging . Mild
cognitive impairment . Alzheimers disease . Dementia .
Vascular

Abbreviations
AD
Alzheimers disease
GCA
global cortical atrophy
MCI
mild cognitive impairment
MMSE mini mental state exam
MTA
medial temporal atrophy
PRD
potential reversible dementia
VAS
visual analogue scale

Introduction
* Martijn V. Verhagen
mverhagen@kg.nl
1

Department of Radiology, Spaarne Gasthuis, Boerhaavelaan 22,

2035 RC Haarlem, The Netherlands

Department of Geriatrics, Spaarne Gasthuis, Boerhaavelaan 22, 2035

RC Haarlem, The Netherlands

Dementia is the overarching term used to describe a wide

range of clinical syndromes consisting of progressive cognitive deterioration and progressive failure to sustain independent living caused by neurodegeneration or damage (e.g.,
Alzheimers disease (AD), vascular dementia, Lewy body dementia, frontotemporal dementia). For 2010, it was estimated

Eur Radiol

that the worldwide prevalence of people over the age of 60

living with AD was 35.6 million [1]. Recent studies commissioned by Alzheimers Disease International (ADI) have estimated that the prevalence of dementia will double between
2001 and 2040 in developed countries and quadruple in undeveloped countries [2].
When a patient is suspected of suffering from dementia,
because of either subjective or objective cognitive deterioration, a geriatrician (or a neurologist or a psychiatrist) is usually
in charge of the clinical workup of the dementia subtype diagnosis. They are able to do this with high accuracy [25], but
will, if necessary, consult other specialists (e.g., neurologist,
psychiatrist, psychologist and/or radiologist) before making a
final dementia subtype diagnosis.
In diagnosing dementia, it is recommended by the European
Federation of Neurological Societies (EFNS) [68], the
National Institute for Health and Care Excellence (NICE)
(UK) [9] and the American Academy of Neurology (AAN)
[10] to perform at least one neuroimaging study such as computed tomography (CT) or magnetic resonance imaging (MRI)
of the brain. Historically, this was advocated to establish potential reversible dementia (PRD; e.g., subdural haematoma,
brain tumour) [11]. However, for MRI of the brain, several
visual rating scales (semi-quantitative human observer
scoring scales) have been developed that are used to specify
the dementia subtype [12]. The best known [13] are Scheltenss
medial temporal lobe atrophy (MTA) [14, 15], global cortical
atrophy (GCA) [16], and white matter hyperintensity according to the Fazekas scale as a measure for small vessel ischaemic
disease [17]. These visual rating scales have shown high specificity [1417] and are commonly applied to neuroimaging
studies. Specifically, the application of the MTA visual rating
scale is recommended in a recent guideline for AD [4]. In
addition to these visual rating scales, it is important to report
regional and asymmetric atrophy (e.g., frontotemporal atrophy,
parietal atrophy) [12, 13]. MRI is also highly sensitive to focal
vascular pathology (e.g., microbleeds, lacunes) [4].
Correct dementia subtype diagnosis may have several implications. The most apparent is the finding of a PRD, because this
may be treated surgically. Furthermore, the specific dementia
subtype diagnosis may have implication for the prognosis, the
prescription of disease-modifying drugs, and psychological and
supportive therapy [9]. For example, the diagnosis of mild cognitive impairment (MCI) has significant consequence for the
patient because a proportion (60-65 %) of this population is
suspected to progress to clinical dementia during their life [18].
Despite the availability of visual rating scales on MRI, and
the potential implication on patient diagnosis and treatment, to
our knowledge, no studies have been published that demonstrate the actual value of visual rating scales on the clinical
workup of dementia. Therefore, our first goal was to establish
in which way MRI, combined with visual rating scales, affects
the clinical diagnosis. The second goal was to gain insight into

the influence of neuroimaging on the confidence level of the

geriatrician in the final diagnosis.

Methods
The study design consisted of a prospective observational single
centre study, performed in a general hospital (Spaarne Gasthuis,
Haarlem, the Netherlands) from October 2013 to July 2014.
Approval was acquired from the local institutional review board.
All patients with cognitive deterioration referred to our department of geriatrics were consecutively enrolled into the
study. The inclusion criterion was first presentation with cognitive deterioration; exclusion criteria were inability to undergo
MRI (e.g., because of a pacemaker or inability to remain still),
neuroimaging from elsewhere performed as part of the workup
of cognitive deterioration, known neurodegenerative disease or
significant brain damage, and/or a known brain tumour.
Patients were evaluated by one of four geriatricians, all with
more than five years of clinical experience with dementia. Depending on the specific needs of the patient, the
geriatrician consulted supporting specialists (e.g., a neurologist, neuropsychologist and/or psychiatrist). A mini
mental state examination (MMSE) [19] and clock drawing test were determined in all patients. Vascular comorbidity, prior cerebral vascular accident (CVA), diabetes,
history of smoking, heightened cholesterol levels, and
blood pressure were retrieved from the patient file.
All patients underwent an MRI examination as part of the
standard workup of cognitive deterioration in our hospital.
The MRI studies were examined as part of a daily workload
by one of five radiologists, all with more than five years of
experience with neuroimaging. Studies were examined and
reported according to standard practice. The visual rating
scales used were; global cortical atrophy (GCA) [16], medial
temporal atrophy (MTA) [14, 15], and white matter
hyperintensity according to the Fazekas scale as a measure
for small vessel ischaemic disease [17]. Regional or asymmetric atrophy not included in the visual rating scales was reported separately, as is part of the normal routine.
During a weekly multidisciplinary meeting the geriatrician
was asked to determine the diagnosis, differential diagnosis,
treatment plan, and level of confidence in the diagnosis by
means of a visual analogue scale (VAS) both prior to, and after
assessing the neuroimaging study accompanied by the radiology report. The MRI study and radiology report were blinded
to the geriatrician until the multidisciplinary meeting.
Only in the case of PRD were the findings immediately
communicated with the geriatrician without waiting for
the weekly meeting. PRD included any potentially reversible or surgically treatable pathology such as brain tumours or
subdural haemorrhage (SDH). The geriatricians final diagnosis was used as the diagnostic standard. The VAS ranged from

Eur Radiol

no confidence to absolute confidence and was subsequently

set to a scale of 0-100.
MRI was performed on a 1.5T scanner (Philips, Amsterdam,
the Netherlands) using a standardized dementia protocol which
included the following sequences: T2-weighted fast spin echo
(FSE; study parameters: transversal and coronal plane, voxel
size 0.9 mm 1.12 mm, slice thickness 3 mm, interslice distance 0.3 mm, TE 100 ms, TR 4643 ms), T2 fluid attenuated
inversion recovery (FLAIR; study parameters: transversal plane,
voxel size 0.9 mm 1.12 mm, slice thickness 4 mm, interslice
distance 0.4 mm, TE 120 ms, TR 6000 ms), 3D T1 fast field
echo (FFE; study parameters: voxel size 0.94 mm 1.13 mm
2 mm, TE 4.6 ms, TR 9.3 ms), DWI (study parameters: transversal plane, voxel size 2.05 mm 2.56 mm, slice thickness 5
mm, interslice distance 1 mm, TE 72 ms, TR 2492 ms).
Additional to these sequences we performed a susceptibilityweighted imaging sequence (SWI; study parameters: transversal
plane, voxel size 0.86 mm 1.15 mm, slice thickness 6 mm,
interslice distance 0.6 mm, TE 48 ms, TR 2584 ms) to detect
possible microbleeds. The imaging studies were stored in the
picture and communication system (PACS) of our hospital.
Statistical analysis
Statistical analysis was performed using SPSS for Windows
(release 22). The level of significance was set at < 0.05. The
confidence interval (CI) for frequencies of change in diagnosis
was calculated by means of bootstrapping; the CI was set at 95
%. 150 consecutive patients were needed, based on a power
analysis with an expected 7 % chance of PRD with a 4 %
range [20]. The Wilcoxon test was used for comparing the
level of confidence determined by VAS before and after neuroimaging. The Kruskal-Wallis non-parametric test was used
to compare levels of confidence with respect to three diagnosis groups, and to compare age and MMSE between dementia
subtype groups. The Mann-Whitney U test was used to compare ordinal variables between dementia subtype groups.
Dichomatous variables between dementia subtype groups
were compared using the Chi-square test.

Results
In the period from October 2013 to July 2014, 148 patients
were enrolled into the study. 131/148 (88.5 %) patients were
referred to a geriatrician by their general practitioner, the remainder were referred by a neurologist. 13 patients were excluded because they could not undergo MRI, these patients
subsequently underwent CT. The population that underwent
MRI had an average age of 75.5 years (median 77, standard
deviation (SD) 9.4, range 45-93), and an average MMSE of
25.6 (median 26, SD 3.2, range 16-30). PRD was reported in
3/135 (2.2 %) patients undergoing MRI: 2 primary brain

tumours, and 1 metastasis of an unknown primary tumour.

No PRD was shown in the patients that underwent CT.
After MRI of the brain, the distribution of the definitive
diagnosis made by the geriatrician consisted of (Table 1): AD
(29.6 %), AD with a vascular component (6.7 %), mild cognitive impairment (MCI; 37.8 %), MCI with a vascular component (6.7 %), depression (6.7 %), PRD (2.2 %), vascular dementia (1.4 %), Lewy body dementia (0.7 %), frontotemporal
dementia (0.7 %) and a rest group (7.4 %) which consisted of
patients without objective cognitive impairment. When combining the groups AD with a vascular component, MCI with a
vascular component, and vascular dementia, 20/135 (14.8 %)
of all patients had either a partial or absolute suspected vascular
aetiology of their cognitive deterioration.
After MRI, the diagnosis changed in 23.7 % (32/135 patients; CI 17.0 % - 31.1 %) of all patients, including PRD
(Table 2). In 63.0 % (CI 54.8 % - 70.4 %), neuroimaging
results did not affect the clinical (differential) diagnosis. In
13.3 % (CI 8.1 % - 19.3 %), the differential diagnosis was
altered, whereas the main diagnosis remained unchanged.
Change in diagnosis was due to a change in suspected vascular aetiology in 18/32 patients (18/135 patients, 13.3 % of all
patients). In 13/18 patients, a vascular component was added
to their main diagnosis or the diagnosis was changed to vascular dementia; 5/18 patients were no longer suspected of a
vascular aetiology after the MRI (Table 3).
Establishing treatment prior to the MRI was not possible in
our setting because, in most patients, the treatment plan was
determined based on the final results of the multidisciplinary
meeting, and in consultation with the patient.
There was a significant overall improvement in the level of
confidence after MRI as expressed by using a VAS (Table 4),
increasing from a mean of 80.9 (SD 15.0) before imaging to a
mean of 87.8 (SD 11.9) after imaging (p = 0.001). Within the
dementia subtypes the mean VAS of both AD (p = 0.001) and
MCI (p = 0.001) improved significantly after MRI.
There was a significantly higher (p < 0.02) Fazekas scale
for the groups MCI with a vascular component, AD with a
vascular component, and vascular dementia when compared
to all other groups, including MCI and AD. Two microbleeds
were found in the MCI population on the SWI sequence, no
microbleeds were found in groups with a vascular component
or vascular dementia (Table 5). Lacunes and other infarctions
were found both in the groups diagnosed with and without a
vascular component (Table 5). No strategic infarcts were reported. The number of focal vascular pathologies was too low
to analyse. MTA was scored significantly higher (p = 0.01) in
patients with AD when compared to any other group. No
significant difference in GCA was found between AD and
MCI; however, there was a significant difference in GCA
between both AD and MCI when compared to the depression
group and the group without objective cognitive impairment
(p = 0.01). No significant differences in regional atrophy were

Eur Radiol
Table 1

Demographics

AD

Before MRI (n) (%)

After MRI (n) (%)

Age (SD)

Gender (% male)

MMSE (SD)

41 (30.4)

40 (29.6)

77.9 (7.9)

35

23.2 (3.6)

AD + vascular component

5 (3.7)

9 (6.7)

82.9 (5.9)

33

23.9 (2.6)

MCI
MCI + vascular component

62 (45.9)
5 (3.7)

51 (37.8)
9 (6.7)

76.3 (6.8)
79.7 (5.9)

49
67

26.7 (2.0)
26.9 (2.1)

Lewy body dementia

Vascular dementia

2 (1.5)
2 (1.5)

1 (0.7)
2 (1.5)

79
84

100
100

23
26

Frontotemporal dementia

1 (0.7)

1 (0.7)

84

100

Depression
PRD

6 (4.4)
3 (2.2)

9 (6.7)
3 (2.2)

62.3 (9.9)
75 (2.8)

44
50

27.8 (1.9)
25.0 (5.7)

8 (5.9)
135

10 (7.4)
135

60.1 (10.5)
75.5 (9.4)

67
47

27.9 (2.3)
25.6 (3.2)

Rest group *
Total
- Missing

* no objective cognitive impairment

after MRI
AD: Alzheimers disease
MCI: Mild cognitive impairment
MMSE: Mini mental state exam
PRD: potential reversible dementia

2004, Condefer [21] already reported a 12 % change in diagnosis after a CT of the brain, however, the study was performed retrospectively, included only CT, and did not use
visual rating scales. We found a 23.7 % change (CI 17.0 % 31.1 %) in diagnosis after MRI in a prospective study including visual rating scales; we confirm and provide stronger evidence for the significant impact of neuroimaging on the final
dementia subtype diagnosis.
Change in diagnosis was due to the detection of a vascular
aetiology in a significant group of patients (13.3 % of all
patients). This confirms that it is difficult to assess vascular
comorbidity based on clinical findings alone. The demonstration of vascular abnormalities has therapeutic implications; for

described. Patients diagnosed with MCI or AD were significantly older than patients in the depression group and the rest
group without cognitive impairment (p = 0.001). MMSE was
significantly lower in AD compared to all other groups (p =
0.001). The clock drawing test was significantly more often
false in the AD group compared to all other groups (p = 0.02).

Discussion
The purpose of this study was to establish the impact of MRI
of the brain, combined with visual rating scales, on the clinical
workup of patients presenting with cognitive deterioration. In

Table 2 Change in diagnosis

after MRI per dementia subtype
(n > 2)
AD
AD + vascular component
MCI
MCI + vascular component
Depression
Rest group *
Overall (%)

Same diagnosis

Different diagnosis

Same diagnosis, different DD

65.0 %
44.4 %
78.4 %
33.3 %
55.6 %
70.0 %
63.0 %

17.5 %
55.6 %
5.9 %
66.7 %
33.3 %
20.0 %
23.7 %

17.5 %
0%
15.7 %
0%
11.1 %
10.0 %
13.3 %

*no objective cognitive impairment

AD: Alzheimers disease
DD: differential diagnosis
MCI: Mild cognitive impairment

Eur Radiol
Table 3

Change in diagnosis

Diagnosis prior to MRI

Number of changes after MRI

AD

AD
AD + vascular component
MCI

AD+vasc

MCI

Total
MCI+vasc

LBD

VD

FTD

Depression

Rest*
8
1

1
5

MCI + vascular component

LBD

14

2
1

VD
FTD

Total

2
1

1
5

29

* no objective cognitive impairment

AD: Alzheimers disease
FTD: frontotemporal dementia
LBD: Lewy body dementia
MCI: Mild cognitive impairment
PRD: potential reversible dementia
VD: vascular dementia

instance, the progression of small vessel ischaemic disease

may be treated by medication. The Fazekas scale was significantly higher in groups diagnosed with a vascular component
than without. Therefore, we conclude that the Fazekas scale
has a significant impact on the diagnosis of a vascular component during the dementia workup. Focal vascular pathologies (lacunes, microbleeds) were seen in groups diagnosed
both with and without a vascular component, though statistical
analysis was not possible because of the low number of focal

findings. The presence of these focal pathologies in both

groups might imply a less important role in the diagnosis of
a vascular component when compared to the Fazekas scale.
We found a 2.2 % incidence of PRD, which is low compared to prior studies which ranged from 2.8 % to 9 % [18, 20,
21]. This difference might be explained by a variation in definition of PRD in these previous studies. For example, Farina
[19] includes all drug, metabolic, and psychiatric causes of
cognitive deterioration (e.g., depression). The low incidence

Table 4 Change in level of

confidence (expressed using a
VAS)

Groups after MRI

Mean VAS before

imaging (SD)

Mean VAS after

imaging (SD)

Mean Change
VAS (SD)

AD
AD + vascular component
MCI
MCI + vascular component
Lewy body dementia ^

78.1 (15.5)
80.3 (10.4)
82.3 (17.1)
84.7 (8.5)
90

86.8 (11.8)
86.3 (14.7)
90.6 (10.8)
89.8 (10.4)
70

+8.7 (9.8)*
+6.0 (6.7)
+8.3 (14.4)*
+5.2 (8.1)
-20

Vascular ^
Frontotemporal ^
Depression
Rest group *
Total

67
85
85.8 (13.4)
80.7 (17.8)
80.9 (15.04)

94
70
88.5 (11.3)
84.9 (13.8)
87.8 (11.85)

+27
-15
+2.8 (3.4)
+4.1 (9.9)
+6.9 (11.5)*

^ n = 1 or 2
* p < 0.05
* no objective cognitive impairment
AD: Alzheimers disease
MCI: Mild cognitive impairment
VAS: Visual Analogue Scale (0-100)

Eur Radiol
Table 5 Vascular pathology
findings

Vascular pathology

AD

AD+vasc

MCI

Number of patients

51

40

MCI+vasc
9

2.22 (1.09)*

3.00 (0^)

Fazekas (scale 0-3; mean, SD)

1.29 (0.91)

2.44 (0.53)*

1.29 (0.91)

Microbleeds
Lacunes

2
3

Old supratentorial infarction

9 (22.5 %)

5 (55.6 %)

Old cerebellar infarction

Total number (%)

5 (9.8 %)

2 (22.2 %)

VD

0 (0 %)

^n=2
* p < 0.05
AD: Alzheimers disease
MCI: Mild cognitive impairment
Vasc: Vascular component
VD: Vascular dementia

of PRD in our population may also be caused by a

selection bias. In the Netherlands, patients with cognitive deterioration are usually referred to a geriatrician,
whereas patients in whom cognitive deterioration is less
apparent will be referred to another medical specialist
(e.g., a neurologist). In this way, patients with PRD
who would have been included in previously published
studies might not have been enrolled in our study.
This is the first study in which the confidence level of the
geriatrician in the dementia subtype diagnosis was determined
before and after MRI of the brain. We showed a significant
overall improvement of the confidence level in the final diagnosis. A major source of the lower confidence level prior to
the MRI might be concern for a PRD. This concern is the main
reason all patients presenting with cognitive deterioration undergo an imaging study. The exclusion of a PRD will, therefore, raise the confidence level of the referring geriatrician.
Another source of lower pre-imaging confidence might be
vascular aetiology, which is known to be more difficult to
diagnose clinically and for which MRI is highly sensitive
[4]. We confirm this by showing a 13.3 % change in diagnosis
in all patients due to the demonstration of a vascular aetiology.
An improved confidence level after imaging may improve
patient-physician communication and help in defining the
treatment plan (e.g., level of strictness in regulating blood
pressure, cholesterol, diabetes); however, further studies on
this matter are necessary.
One of the limitations of our study was the lack of double
reading of the MRI studies. Our intention was to demonstrate
the influence of MRI of the brain in daily practice. In this
setting, double reading was not achievable due to the high
work flow in the daily routine. Our findings concerning the
distribution of GCA, MTA, and the Fazekas scale between the
dementia subtypes are in accordance with double reading
studies [1417, 22]. As interobserver reliability of the visual

rating scales are known to be moderate, we do not expect a

significant interobserver bias in our study [1417, 22].
Though the ultimate confirmation for diagnosing dementia
is pathology, for obvious reasons, most studies that investigate
dementia lack pathology. We accepted the geriatricians final
diagnosis as the diagnostic standard. Larson [3] showed a diagnostic accuracy of 92 % in AD comparing clinical diagnosis
with autopsy; however, the specificity of other dementia is lower and varies greatly (23-88 %) [4]. This implies that the diagnostic standard we used is suboptimal for dementias other than
AD. In these patients, we cannot be certain that MRI improves
the accuracy of the dementia subtype diagnoses. Further research with a neuropathological reference standard is needed.
With an expected increase in dementia prevalence, the use
of MRI in dementia analysis will increase. This study shows
that MRI of the brain, combined with visual rating scales, goes
beyond excluding PRD, and MRI has a significant impact on
the dementia subtype diagnosis in clinical practice and, additionally, on the geriatricians confidence in his/her final diagnosis. Furthermore, this study confirms the importance of
MRI in demonstrating vascular disease as a cause of cognitive
deterioration. Finally, we conclude that every patient with
cognitive deterioration should undergo an MRI of the brain
in combination with visual rating scales.
Acknowledgments We thank Ralph Vreeswijk (Msc, department of
Geriatrics, Spaarne Gasthuis, Haarlem, the Netherlands), for his help with
the enrolment of the patients. The scientific guarantor of this publication
is Gerard L. Guit (MD, PhD, radiologist, Spaarne Gasthuis, Haarlem, the
Netherlands). The authors of this manuscript declare no relationships with
any companies whose products or services may be related to the subject
matter of the article. The authors state that this work has not received any
funding. Tjeerd van der Ploeg (MsC, statistician, Spaarne Gasthuis, Haarlem, the Netherlands) kindly provided statistical advice for this manuscript. Institutional review board approval was obtained. Written informed consent was waived by the institutional review board. Methodology: prospective, observational, performed at one institution.

Eur Radiol

References
1.

2.
3.

4.

5.

6.

7.

8.

9.

10.

Prince M, Bryce R, Albanese E, Wimo A, Ribeiro W, Ferri CP

(2013) The global prevalence of dementia: A systematic review
and metaanalysis. Alzheimers Dement 9:6375
Ferri CP, Prince M, Brayne C et al (2005) Global prevalence of
dementia: a Delphi consensus study. Lancet 9503:21122117
Larson EB, Edwards JK, O'Meara E, Nochlin D, Sumi SM
(1996) Neuropathologic diagnostic outcomes from a cohort
of outpatients with suspected dementia. J Gerontol A Biol
Sci Med Sci 6:313318
Dubois B, Feldman HH (2007) Research criteria for the diagnosis
of Alzheimer's disease: revising the NINCDSADRDA criteria.
Lancet Neurol 8:734746
Sorbi S, Hort J, Erkinjuntti T et al (2012) EFNS-ENS Guidelines on
the diagnosis and management of disorders associated with dementia. Eur J Neurol 9:11591179
Filippi M, Agosta F, Barkhof F et al (2012) EFNS task force: the
use of neuroimaging in the diagnosis of dementia. Eur J Neurol 19:
14871501
Waldemar G, Dubois B, Emre M et al (2007) Recommendations for
the diagnosis and management of Alzheimer's disease and other
disorders associated with dementia: EFNS guideline. Eur J Neurol
14:126
Hort J, O'Brien JT, Gainotti G et al (2010) EFNS guidelines for the
diagnosis and management of Alzheimer's disease. Eur J Neurol 17:
12361248
NICE (2011) Dementia: Supporting people with dementia and their
carers in health and social care. Available via https://www.nice.org.
uk/guidance/cg42/evidence/cg42-dementia-full-guidelineincluding-appendices-172
Knopman DS, DeKosky ST, Cummings JL et al (2001) Practice
parameter: diagnosis of dementia (an evidence-based review).
Report of the Quality Standards Subcommittee of the American
Academy of Neurology. Neurology 56:11431153

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.
21.

22.

Chui H, Zhang Q (1997) Evaluation of dementia: a systematic study

of the usefulness of the American Academy of Neurology's practice
parameters. Neurology 4:925935
Harper L, Barkhof F, Scheltens P, Schott JM, Fox NC (2014) An
algorithmic approach to structural imaging in dementia. J Neurol
Neurosurg Psychiatry 6:692698
Harper L, Barkhof F, Fox NC, Schott JM (2015) Using visual rating
to diagnose dementia: a critical evaluation of MRI atrophy scales. J
Neurol Neurosurg Psychiatry. doi:10.1136/jnnp-2014-310090
Scheltens P, Fox N, Barkhof F, De Carli C (2002) Structural magnetic resonance imaging in the practical assessment of dementia.
Lancet Neurol 1:1321
Scheltens P, Launer LJ, Barkhof F, Weinstein HC, van Gool WA
(1995) Visual assessment of medial temporal lobe atrophy on magnetic
resonance imaging: interobserver reliability. J Neurol 242:557560
Scheltens P, Pasquier F, Weerts JG, Barkhof F, Leys D (1997)
Qualitative assessment of cerebral atrophy on MRI: inter- and
intra-observer reproducibility in dementia and normal ageing. Eur
Neurol 37:9599
Wahlund LO, Barkhof F, Fazekas F et al (2001) A New Rating
Scale for Age-Related White Matter Changes Applicable to MRI
and CT. Stroke 32:13181322
Busse A, Angermeyer MC, Riedel-Heller SG (2006) Progression of
mild cognitive impairment to dementia: a challenge to current
thinking. Br J Psychiatry 5:399404
Folstein MF, Folstein SE, McHugh PR (1975) Mini-mental state".
A practical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res 3:189198
Farina E, Pomati S, Mariani C (1999) Observations on dementias
with possibly reversible symptoms. Aging (Milano) 5:323328
Condefer KA, Haworth J, Wilcock GK (2004) Clinical utility of
computed tomography in the assessment of dementia: a memory
clinic study. Int J Geriatr Psychiatry 19:414421
Wardlaw JM, Ferguson KJ, Graham C (2004) White matter
hyperintensities and rating scales-observer reliability varies with
lesion load. J Neurol 5:584