Structural Ultrasound of the Medial Temporal Lobe in Alzheimer’s Disease

Abstract

Purpose One of the anatomical hallmarks of Alzheimer’s disease (AD) is the atrophy of the medial temporal lobe (MTL), yet cost-effective and broadly available methodological alternatives to the current imaging tools for screening of this brain area are not currently available.

Materials and Methods Using structural transcranial ultrasound (TCS), we attempted to visualize and measure the MTL, and compared the results of 32 AD patients and 84 healthy controls (HC). The MTL and the surrounding space were defined in the coronal plane on TCS. A ratio of the height of the MTL/height of the choroidal fissure (M/F) was calculated in order to obtain a regional proportion.

Results An insufficient temporal bone window was identified in 22 % of the AD patients and 12 % of the HCs. The results showed that the ratio of M/F was significantly smaller in the AD group on both sides (p = 0.004 right, p = 0.007 left side). Furthermore, the M/F ratio made it possible to discriminate AD patients from HCs with a sensitivity of 83 % (right)/73 % (left) and a specificity of 76 % (right)/72 % (left) which is basically comparable to results published for magnetic resonance imaging. The measurements showed substantial intra/interrater reliability (ICC:0.79/0.69).

Conclusion These results suggest that utilization of structural TCS may possibly constitute a cheap and easy-to-use supplement to other techniques for the diagnosis of AD. It may be especially useful as a screening tool in the large population of individuals with cognitive decline. Further studies are needed to validate this novel method.

Zusammenfassung

Ziel Eines der charakteristischen anatomischen Kennzeichen der Alzheimer-Demenz (AD) ist die Atrophie des medialen Temporallappens (MTL), jedoch fehlen bisher preiswerte und breit verfügbare Alternativen zu den üblichen Bildgebungsverfahren um Veränderungen dieser Gehirnregion darzustellen.

Material und Methoden Mittels transkranieller B-Bild-Sonografie (TCS) haben wir den MTL dargestellt, vermessen und die Messungen von 32 AD-Patienten und 84 gesunder Kontrollen miteinander verglichen. Der MTL und die umgebenden Strukturen wurden zunächst in der koronaren Ebene in der TCS identifiziert. Der Quotient aus der Höhe des MTL und der Höhe der choroidalen Fissur (M/F) wurde als regionales Maß berechnet.

Ergebnisse Bei 22 % der AD-Patienten und bei 12 % der gesunden Kontrollen lag ein insuffizientes Knochenfenster vor. Der M/F-Quotient ergab auf beiden Seiten signifikant kleinere Werte in der AD-Gruppe (p = 0,004 rechte Seite, p = 0,007 linke Seite). Weiterhin konnten AD-Patienten von gesunden Kontrollen mittels des M/F-Quotienten mit einer Sensitivität von 83 % (rechts) / 73 % (links) und einer Spezifität von 76 % (rechts) / 72 % (links) von einander differenziert werden, was mit publizierten Werten für die Magnetresonanztomografie vergleichbar ist. Es ergaben sich zudem eine substanzielle Intra- und Interrater-Reliabilität (ICC:0,79/0,69).

Schlussfolgerung Unsere Ergebnisse legen die Nutzung der TCS als günstige und einfach anzuwendende Ergänzung zu anderen Verfahren in der Diagnostik von AD nahe. TCS könnte insbesondere als Screening-Instrument in großen Kollektiven von Menschen mit kognitiven Defiziten sinnvoll sein. Weitere Studien sind notwendig, um diese neue Methode zu validieren.

Key words

transcranial sonography - medial temporal lobe - Alzheimer's disease

Introduction

When diagnosing Alzheimer’s disease (AD), the most prominent anatomical region for identifying disease-related changes is the medial temporal lobe (MTL). The detection of atrophy in the MTL primarily using magnetic resonance imaging (MRI) has been a subject of research for decades [1]. Various MRI-associated strategies such as visual assessment scales (VAS) [2] [3], manual or automated assessments [4], or hippocampal subfield measurements [5] have been employed to determine the extent of MTL atrophy. It has repeatedly been shown that changes in the MTL can differentiate patients with AD or mild cognitive impairment (MCI) from healthy controls (HCs) [6], correlate with cognitive decline [7], and predict AD in individuals with MCI [8].

Sonographic technologies are currently primarily limited to the vascular field in dementia diagnostics. As vascular pathologies increase the risk and the progression of AD and as there is a huge overlap between AD and vascular dementia [9], evaluation of a suspected patient with Doppler ultrasound often provides valuable additional information for clarification of the diagnosis and underlying pathophysiology [10].

In the field of dementia, TCS is rarely utilized apart from the vascular domain. Structural TCS has previously been applied only to show enlargement of the ventricles in demented patients [11], or in the context of non-motor features of Parkinson’s disease (PD) [12]. Assessment of the hippocampal area with TCS has not been reported, so far. In the present study, we aimed to determine whether the anatomical quantification of MTL is feasible with structural TCS, and whether it could be used for the differentiation of patients with AD from HCs.

Methods and Materials

Patients

AD patients were recruited from the clinics of the Departments of Neurology and Psychiatry at the University of Tuebingen between December 2014 and November 2015. All patients fulfilled the criteria for “probable AD with increased level of certainty” according to the latest diagnostic guideline [13]. HCs were recruited from the ongoing third follow-up of the “Tübinger Evaluation of Risk Factors for Early Detection of Neurodegeneration” (TREND) study, in which elderly individuals are invited biennially for quantitative assessment of symptoms supposed to precede the diagnosis of PD or AD. Since the TREND cohort is highly selective for these markers, strict criteria were applied to rule out individuals with an already ongoing neurodegenerative process. Hence, participants with MCI, moderate or severe depressive state on Beck’s Depression Inventory (BDI), and hyposmia were excluded from the HC group.

The study was approved by the ethical committee of the Medical Faculty of the University of Tuebingen. All procedures were in accordance with the Declaration of Helsinki. Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Tuebingen[14].

Sonographic image acquisition

Ultrasound settings and assessment of ventricles in the axial plane

All sonographic measurements were performed by the same examiner (RY). Assessment of ventricles was performed using a phased-array ultrasound probe with a transducer frequency of 2.5 MHz, and with an axial and lateral resolution of approximately 0.7, 3 mm, respectively (Esaote MyLab Alpha, Genoa, Italy). Examination was performed through the preauricular acoustic bone window of the temporal bone with a penetration depth of 16 cm and a dynamic range of 45 dB. The assessment was achieved in the axial plane at the level of the “orbitomeatal line” which allows visualization of the base of the cranium. By tilting the ultrasound probe about 10° upwards, the diencephalic plane was visualized in a slightly oblique section. On this plane the anechogenic third ventricle and the frontal horn of the contralateral ventricle were identified. Measurements of ventricles were achieved in the frozen and zoomed image. For details of the measurements in the mesencephalic and ventricular planes, see [15].

Assessment of the MTL in the coronal plane

Visualization of the MTL was carried out with the same transducer. The frequency of the transducer was increased to up to 4 MHz, and the focal zone was adjusted appropriately in order to improve the lateral resolution. Holding the probe vertical to the orbitomeatal line ([Fig. 1a]) allowed visualization of the hypoechogenic MTL and the surrounding hyperechogenic cerebrospinal fluid space in the 2-dimensional coronal plane ([Fig. 1b]). After defining the structures, measurements were performed ipsilaterally after freezing and zooming of the respective image. The brain stem adjacent to the MTL was taken as a landmark. The structures were measured only in planes in which the MTL was located adjacent to the anterior or mid-brain stem. An examination was defined as “insufficient” when the image quality was not adequate for reliable measurements due to the lack of a sufficient acoustic window of the temporal bone. The term “poorly sufficient” was used when only some of the measurements could be obtained.

Fig. 1 Coronal image of a healthy brain of a participant in MRI a and TCS b in the supine position. a) The probe is placed vertical to the middle temporal bone to obtain the coronal image. The brain stem, ventricles and ipsilateral MTL are visualized. b) Coronal image obtained by TCS. The lateral ventricles (L), third ventricle (T) and structures surrounding the MTL are depicted. White arrow: choroidal fissure; empty arrow: ambient cistern; arrow head: temporal horn of the lateral ventricle.

Measurements of the MTL

After a pilot assessment of 20 individuals (which is not included in these analyses), the best image depicting the MTL and the surrounding area was identified, followed by the assignment of the following five linear measurements ([Fig. 2]).

Fig. 2 Linear measurements of MTL on MRI a and TCS b in a healthy participant on the coronal plane.

A: Horizontal line from the temporal horn of the lateral ventricle to the bottom of the temporal lobe,

B: Horizontal line from the midpoint of the MTL to the bottom of the temporal lobe (height of the MTL),

C: Vertical line from the temporal horn of the lateral ventricle to the ambient cistern (width of the MTL),

D: Vertical line of the extension of the ambient cistern (width of the ambient cistern),

E: Horizontal line measuring the extension of the choroidal fissure in the middle of this structure (height of the choroidal fissure).

As the total brain size of the cohort was not assessed, the influence of the individual variability of the total brain size on the MTL measurements could not be accounted for. Therefore, a ratio of MTL height over choroidal fissure height (B/E) was also calculated in order to avoid this confounding effect, and to obtain a relation within the MTL. These two (B and E) measurements appeared to be most relevant, as the heights of the hippocampus and choroidal fissure are negatively correlated in the process of MTL atrophy.

Statistical analyses

Descriptive and quantitative data are given as mean and SD. Intra-class correlations (ICC) were assessed by re-measuring the images under blinded conditions with a one to three-month interval by the sonographer and by a second rater blinded to the initial measurements. Data of the two groups was compared using a chi-square test, student’s t-test or Mann-Whitney U test as appropriate. Logistic regression was further performed to assess the predictive characteristics of the MTL measurements between HCs and the AD group, with age, sex and ventricle measurements included as covariates. Receiver operating characteristics (ROC) curves were analyzed in order to define a cut-off value for the highest sensitivity and specificity of the technique.

Results

Of a total of 202 HCs, 52 were excluded due to MCI, 33 due to hyposmia, 24 (12 %) due to insufficient bone window on both sides, and 9 were excluded due to depression, resulting in 84 HCs eligible for further analyses. In the case of only one-sided insufficiency of the temporal bone window, participants were included with the one side on which measurements could be performed. This was the case in 24 participants of the HC group. Out of the 41 patients with AD, 9 (22 %) were excluded due to an insufficient bone window on both sides on TCS, and in the remaining 32 patients, 8 had one-sided insufficiency. Thus a total of 72 (17.8 %) of all 404 examinations (one on each side) of the 202 HCs and 26 (31.7 %) of 82 examinations of 41 AD patients could not be performed due to an insufficient bone window and were therefore not included in the statistics. The one-sided insufficiencies caused missing data on either side which are shown in detail in table 2. Moreover, not insufficient but “poorly sufficient” examination conditions were also detected in 20 assessments of the HCs and 15 of the AD group which prevented accomplishment of all of the measurements in some cases, resulting in the missing data.

In the AD group, 8 patients had mild, 22 had moderate, and 2 had severe dementia according to the clinical dementia rating scale. The AD group was significantly older than the HCs and had (expectedly) a lower Mini-Mental State Examination (MMSE) score. Groups were similar with regard to sex, and ventricle sizes were significantly larger in the AD group ([Table 1]) ([Fig. 3]).

Table 1 Comparison of the demographic characteristics and the ventricle enlargements between groups.

	controls mean (SD)	AD mean (SD)	P-value
age	67.9 (6.8)	72.9 (11.1)	0.02[1] [2]
sex, male %	47.6	37.5	0.33[3]
MMSE	28.7 (1.2)	17.1 (6.2)	< 0.0001[4] 2
third ventricle mm.	6.1 (1.7)	8.6 (2.3)	< 0.00011 2
lateral ventricle, right mm.	17.4 (2.3)	19.5 (2.6)	0.00011 2
lateral ventricle, left mm.	18.0 (1.9)	19.3(2.0)	0.0051 2

¹ Student’s t-test.

² p-value significant.

³ Chi-square test.

⁴ Mann-Whitney U test.

Fig. 3 Comparison of TCS images of a control subject a and an AD patient b. Arrows indicate the height of the MTL (long) and the choroidal fissure (short) in each image. Note the differences in the widths of the third (T) and the lateral (L) ventricles.

The measurements selected for the assessment of the MTL and the missing data are shown in table 2. The student’s t-test showed that the measurements of A and B were significantly smaller in the AD group on the right side. Moreover, the size (height) of the choroidal fissure (E) was significantly larger in the AD group on both sides. The width of the MTL (C) and the ambient cistern (D) were not significantly different between groups.

The comparison of the B/E ratios of the groups showed that the HCs had significantly larger values on both sides ([Table 2]) ([Fig. 4]). These ratios correlated significantly with MMSE scores on both sides when the two groups were combined (right side: r = 0.39, p < 0.0001; left side: r = 0.33, p = 0.002). For the intra-rater and inter-rater effects for B/E ratios, substantial agreement was found. The single measure ICC was 0.79 (95 % CI: 0.69 – 0.86) for intra-rater reliability and 0.69 (95 % CI: 0.50 – 0.82) between two raters.

Table 2 Comparison of the measurements between groups.

	controls (n = 84)					AD (n = 32)			P-value[1]
	N		Missing n (%)	mean mm (SD)	N		Missing n (%)	mean mm (SD)
A right		57	27 (32)	12.9 (2.2)	20		12 (38)	10.8 (2.3)	0.001[2]
A left		59	25 (30)	11.9 (2.2)	18		14 (44)	11.4 (1.9)	0.31
B right		67	17 (20)	14.7 (2.1)	24		8 (25)	13 (2.4)	0.0022
B left		68	16 (19)	14.1 (2.4)	26		6 (19)	13.5 (2.3)	0.25
C right		59	25 (30)	16.5 (2.0)	19		13 (41)	15.8 (2.1)	0.19
C left		61	23 (27)	15.8 (2.1)	20		12 (38)	15.4 (2.6)	0.45
D right		71	13 (15)	5.6 (1.2)	21		11 (34)	5.8 (1.4)	0.47
D left		63	21 (25)	5.6 (1.4)	22		10 (31)	5.8 (1.4)	0.44
E right		67	17 (20)	5.3 (0.9)	24		8 (25)	6.3 (1.4)	0.0042
E left		68	16 (19)	5.2 (0.9)	24		8 (25)	6.2 (1.0)	< 0.00012
B/E right[3]		67	17 (20)	2.82 (0.6)	23		9 (28)	2.11 (0.6)	< 0.00012
B/E left3		68	16 (19)	2.83 (0.7)	22		10 (31)	2.22 (0.5)	< 0.0012

mm: millimeter; SD: standard deviation; AD: Alzheimer’s disease.

¹ Student’s t-test.

² p-value significant.

³ Ratios are not depicted with mm.

Fig. 4 Comparison of the MTL images between a healthy control a and an AD patient b. Arrows indicate the height of the MTL (B) and the choroidal fissure (E).

To assess the predictive power of B/E ratios for the likelihood that participants have AD, logistic regressions were performed accounting for age, sex and ventricle widths as covariates. For the right side, the regression model was statistically significant (chi-square(6) = 34.652, p < 0.0001). The model explained 52 % (Nagelkerke R²) of the variance in group membership and correctly classified 84 % of cases. A decrease in the B/E ratio was significantly associated with AD regardless of age, sex and other atrophy measures (ventricle widths) (p = 0.004). For the left side, the overall regression model was also significant (chi-square(6) = 35.482, p < 0.0001), with a strong relationship (Nagelkerke R² = 52 %) between the predictors and grouping, correctly classifying 86 % of the cases. The B/E ratio significantly predicted the AD group while accounting for the confounding variables (p = 0.007).

To define an optimal cut-off value for group classification, ROC analyses were performed for both sides. On the right side a B/E ratio≤ 2.48 separated two groups with a sensitivity of 83 % and a specificity of 76 %, with an area under the curve (AUC) of 0.81 (p < 0.00 001), which corresponds to positive and negative likelihood ratios (LR) of 3.46, 0.22, respectively. On the left side, a ratio of ≤ 2.44 yielded a sensitivity of 73 % and a specificity of 72 % (AUC of 0.76; p < 0.0003), with LR+ = 2.61, and LR– = 0.38.

Discussion

In this study we could show for the first time that it is indeed possible to measure the MTL with TCS. Moreover, the comparison of anatomical measures of structures of this area showed significant differences between AD patients and HCs.

In the pathological process of AD, two of the most prominent changes in the brain are hippocampus atrophy and enlargement of the choroidal fissure which have been the primary focus of VAS or two-dimensional evaluation methods [3]. Especially MRI-based VAS according to Scheltens [2] proved to be a simple and reliable method and applicable in the clinical setting. The results of our analyses are in line with these previous reports that linear measurements of the hippocampal area and the surrounding cerebrovascular space on the coronal plane may help to differentiate AD patients from HCs.

The sensitivity and specificity of MRI assessment of MTL atrophy in the differentiation of AD patients and HCs were reported to be 75 % (95 % CI: 73 – 77) and 81 % (95 % CI: 79 – 82), respectively [16]. In our study, the sensitivity and specificity of TCS for a differentiating cut-off value were identified as 83 %, 76 %, respectively, for the right side and 73 %, 72 %, respectively, for the left side. According to these values, one can argue that structural imaging of the MTL with TCS may be as sensitive as MRI for the identification of MTL atrophy. However, the specificity of TCS was lower compared to MRI. The lower specificity or the overlap between the groups could be attributed to diagnostic error or to the limitations of the technique, such as an inadequate lateral resolution or slight shifts in the examination plane, since the examination planes on TCS could not be defined as exactly as on MRI. This limitation could at least partially be eliminated by using the Fusion Imaging technique with a Virtual Navigator in future studies [17].

The underlying neurodegenerative process in HCs could also play a role in the lower specificity. We have applied strict risk factor exclusion criteria in order to eliminate such controls. However, HCs who are grouped in the AD group by the cut-off value could still be compensating the ongoing degeneration with their cognitive reserve but may develop MCI and AD in the future, as atrophy of the hippocampal area may precede even the MCI stage [18].

Several shortcomings of the study surely need to be addressed. In the AD group, approximately one-third of the sonographic examinations could not be performed due to an inadequate acoustic window which prevents ultrasound waves from passing through the temporal bone. It has been reported that in 10 – 20 % of the normal population, no or only partial visualization of the intracranial structures are possible [19]. Our results in HCs are similar to the previously published reports. The higher percentage of bone window insufficiency in the AD group could be attributed to the higher mean age and/or to the higher percentage of females in whom osteoporosis – which is supposed to be related to a poorer bone window as a result of distraction of ultrasound waves – could be more prevalent. Another point worth discussing is the difference between right and left side measurements. The measurements of “A” and “B” were significantly different between the groups only on the right side. This could be attributed either to the right-handed sonographer or to the structural differences of the hippocampal area. Hippocampal asymmetry between the left and the right side has been previously reported [20], but more studies are needed to evaluate whether such structural differences may be depicted by TCS. Third, the HCs were not age-matched to the AD group which had a significantly higher age (p = 0.02). However, age has been included in the regression model as a predicting variable together with sex and ventricle measurements, and thus MTL differences on both sides should not be confounded by age. Fourth, the examiner who performed the TCS was not blinded to the participants’ diagnostic status. In the clinical setting, perfect blinding of investigators to the participants’ diagnosis is difficult especially when working with AD, since TCS requires direct interaction with the patient (asking the patient to lie in a specific position, to turn the head as needed, etc.). Although the second measurements for inter-rater reliability were performed blinded, one cannot completely rule out an “assessment bias”. Finally, TCS results could not be compared with the MRIs of the patients. Patients with AD in this study had different disease durations. Therefore, MRI had already been performed in the past and could not be used for the validation of the current TCS measurements. Validation of TCS by MRI or the diagnostic gold standard postmortem examination should be done in future studies.

Currently, MRI is one of the most widespread and thus the most frequently used imaging techniques within the field of dementia diagnostics. Structural MRI allows evaluation of the severity and the pattern of the neuronal loss [6] and is relatively cheap compared to functional imaging methods. However, in clinical practice, MRI has its own limitations. It can, for example, not be performed under certain clinical conditions like in individuals with specific cardiac or cochlear implantations, aneurysm clips, prosthesis, shunts, infusion pumps, or claustrophobia. Moreover, MRI is still an expensive method and is time-consuming regarding image acquisition. In comparison, TCS is a cheap, fast and patient-friendly method, which can be performed even at the bedside. It can be repeated quickly as often as required, can be applied within the physical examination session, and the result can be obtained instantly. Although structural TCS is far from being employed for the differential diagnosis of dementia, assessment of the MTL can be performed in patients who have a sufficient bone window but are not suitable for MRI. Furthermore, since a combination of markers generally has a better predictive value than evaluation of single ones [21], MTL evaluation with structural TCS could be combined with other markers of AD for an additive predictive power in the future. Also, a combination with Doppler sonography which could increase the value of sonographic examination in vascular dementia could be considered for future studies.

The search for inexpensive and noninvasive strategies in the field of dementia has always been an important topic, especially in light of the increasing prevalence of AD and other forms of dementia [22]. As far as we are aware, this is the first attempt to employ structural sonographic imaging for the evaluation of the MTL. Taking our results into consideration, we cautiously propose that the height of the MTL divided by the height of the choroidal fissure assessed by TCS may differentiate patients with AD from HCs. These first results may open the diagnostic window for a novel and easy to apply technique in patients with cognitive decline or dementia. Before this method can be applied for following the progression of atrophy or as an early marker for dementia, further research in larger cohorts of AD patients and HCs is needed to validate and confirm this new approach for the assessment of the hippocampal area.

Erratum 20.07.2016: Structural Ultrasound of the Medial Temporal Lobe in Alzheimer’s Disease Yilmaz R, Pilotto A, Roeben B et al. Ultraschall in Med 2016; DOI: 10.1055/s-0042-107150

Unfortunately, the surname of one of the authors was not correctly written and the surname was changed to „Preische“.

• References

• 1 Kehoe EG. McNulty JP. Mullins PG. et al. Advances in MRI biomarkers for the diagnosis of Alzheimer's disease. Biomark Med 2014; 8: 1151-1169
• 2 Scheltens P. Leys D. Barkhof F. et al. Atrophy of medial temporal lobes on MRI in "probable" Alzheimer's disease and normal ageing: diagnostic value and neuropsychological correlates. J Neurol Neurosurg Psychiatry 1992; 55: 967-972
• 3 Kaneko T. Kaneko K. Matsushita M. et al. New visual rating system for medial temporal lobe atrophy: a simple diagnostic tool for routine examinations. Psychogeriatrics 2012; 12: 88-92
• 4 Leung KK. Barnes J. Ridgway GR. et al. Automated cross-sectional and longitudinal hippocampal volume measurement in mild cognitive impairment and Alzheimer's disease. Neuroimage 2010; 51: 1345-1359
• 5 de Flores R. La Joie R. Chetelat G. Structural imaging of hippocampal subfields in healthy aging and Alzheimer's disease. Neuroscience 2015; 309: 29-50
• 6 Teipel S. Drzezga A. Grothe MJ. et al. Multimodal imaging in Alzheimer's disease: validity and usefulness for early detection. Lancet Neurol 2015; 14: 1037-1053
• 7 Jack Jr CR. Petersen RC. Xu Y. et al. Rates of hippocampal atrophy correlate with change in clinical status in aging and AD. Neurology 2000; 55: 484-489
• 8 Devanand DP. Pradhaban G. Liu X. et al. Hippocampal and entorhinal atrophy in mild cognitive impairment: prediction of Alzheimer disease. Neurology 2007; 68: 828-836
• 9 Kalaria R. Similarities between Alzheimer's disease and vascular dementia. J Neurol Sci 2002; 203–204: 29-34
• 10 Keage HA. Churches OF. Kohler M. et al. Cerebrovascular function in aging and dementia: a systematic review of transcranial Doppler studies. Dement Geriatr Cogn Dis Extra 2012; 2: 258-270
• 11 Lauckaite K. Rastenyte D. Surkiene D. et al. Specificity of transcranial sonography in parkinson spectrum disorders in comparison to degenerative cognitive syndromes. BMC Neurol 2012; 12: 12
• 12 Walter U. Skoloudik D. Berg D. Transcranial sonography findings related to non-motor features of Parkinson's disease. J Neurol Sci 2010; 289: 123-127
• 13 McKhann GM. Knopman DS. Chertkow H. et al. The diagnosis of dementia due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011; 7: 263-269
• 14 Harris PA. Taylor R. Thielke R. et al. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009; 42: 377-381
• 15 Berg D. Behnke S. Walter U. Application of transcranial sonography in extrapyramidal disorders: updated recommendations. Ultraschall in Med 2006; 27: 12-19
• 16 Frisoni GB. Bocchetta M. Chetelat G. et al. Imaging markers for Alzheimer disease: which vs how. Neurology 2013; 81: 487-500
• 17 Walter U. Skoloudik D. Transcranial sonography (TCS) of brain parenchyma in movement disorders: quality standards, diagnostic applications and novel technologies. Ultraschall in Med 2014; 35: 322-331
• 18 Smith CD. Chebrolu H. Wekstein DR. et al. Brain structural alterations before mild cognitive impairment. Neurology 2007; 68: 1268-1273
• 19 Berg D. Godau J. Walter U. Transcranial sonography in movement disorders. Lancet Neurol 2008; 7: 1044-1055
• 20 Shi F. Liu B. Zhou Y. et al. Hippocampal volume and asymmetry in mild cognitive impairment and Alzheimer's disease: Meta-analyses of MRI studies. Hippocampus 2009; 19: 1055-1064
• 21 Westman E. Muehlboeck JS. Simmons A. Combining MRI and CSF measures for classification of Alzheimer's disease and prediction of mild cognitive impairment conversion. Neuroimage 2012; 62: 229-238
• 22 Handels RL. Wolfs CA. Aalten P. et al. Diagnosing Alzheimer's disease: a systematic review of economic evaluations. Alzheimers Dement 2014; 10: 225-237