Online newspaper of professor Yasser Metwally

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

Neuroimaging has become increasingly important in the clinical assessment and diagnosis of dementia. Structural imaging with MRI and functional imaging techniques, such as positron emission tomography and single photon emission CT, increasingly are used to aid in the differential diagnosis and early detection of dementia. Imaging techniques also can track disease progression over time and may be useful to monitor treatment effects. The most important development in the field over the past decade is the ability to image amyloid in the brain. This technique will revolutionize patient management and care.

• Structural neuroimaging

The traditional role of structural imaging is to identify structural and potentially treatable causes of cognitive impairment, for example brain tumors, hydrocephalus, and subdural haematomas. X-ray CT and MRI scanning detect most of these pathologies. Structural imaging also increasingly has been recognized as an important tool in the diagnosis of degenerative dementias. A CT or MRI scan has become a routine part of the clinical workup to aid in differential diagnosis. Recommendations from the American Academy of Neurology are that at least one structural scan be performed in all patients who have dementia. CT scans are widely available, cheap, relatively rapid, and appropriate for patients who present acutely with impaired cognition or an acute decline from a previous level. The improved tissue contrast and the ability to detect focal temporal lobe abnormalities, however, mean MRI has several advantages in evaluating brain structures and is more appropriate for nonemergent evaluation. MRI also avoids ionizing radiation. Hence, MRI is one of the most widely used imaging techniques in the assessment of the degenerative dementias.

Different patterns of atrophy can be identified on visual inspection of MRI in different neurodegenerative conditions. Alzheimer’s disease (AD) is the most common neurodegenerative disorder, affecting approximately 4.5 million people in the United States, and, therefore, has received the most attention. Patients who have AD often show patterns of cerebral atrophy involving the medial temporal lobe, in particular the hippocampus and entorhinal cortex, and the posterior cingulate, precuneus, and the temporoparietal association neocortex, with concurrent expansion of the ventricles. There is a relative sparing of the sensorimotor cortex, visual cortex, and cerebellum. The first structural changes seem to occur in the medial temporal lobe, with early volume loss of the hippocampus and entorhinal cortex (Fig. 1). This matches the progression of neurofibrillary pathology in AD, which starts in the entorhinal cortex and medial temporal lobe, before spreading to other limbic areas and out into the neocortex. Visual qualitative rating scales have been developed for the medial temporal lobe, which are quick and easy to apply in clinical practice. These assessments can differentiate patients who have AD from controls with a sensitivity ranging from 40% to 95% and a specificity of 90% [1]. More specific measurements of medial temporal structures also have been performed, particularly of the hippocampus and entorhinal cortex, which again have shown good separation of patients who have AD from controls. Hippocampal atrophy is shown to be a sensitive marker to pathologic AD stage and consequent cognitive status [2]. Medial temporal lobe atrophy is not specific to AD, however, and is seen commonly in other dementias, which limits its usefulness for differential diagnosis of AD. In addition, although the presence of medial temporal atrophy makes the diagnosis of AD more likely, the absence of atrophy does not exclude the diagnosis.

Automated assessment techniques also have been used recently to investigate patterns of cerebral atrophy on MRI in AD and other dementias. One of the most common techniques is voxel-based morphometry (VBM), which compares groups of patients and identifies differences in the patterns of cerebral atrophy across the whole brain. This has advantages over region-of-interest based techniques in that it does not require any a priori decisions concerning which structures to assess and can provide more detailed information about cortical changes. VBM studies show that although the regions of greatest loss occur most often in the medial temporal lobe, there also is extensive atrophy throughout the temporal lobe, parietal lobe, posterior cingulate and precuneus, insula, temporoparietal association neocortex, and prefrontal gyri in subjects who have a clinical diagnosis of AD compared with controls (Fig. 2) [3]. Structures in the central gray matter also are involved, including the caudate, putamen, thalamus, and hypothalamus. VBM studies show a difference in the patterns of atrophy in patients who have an early (=65 years) versus late (>65-year) disease onset. Patients who have late onset tend to show a pattern of loss relatively restricted to the medial temporal lobes, whereas patients who have early onset show a more widespread pattern of atrophy involving the temporoparietal association neocortex, precuneus, and frontal lobes [4]. This technique, therefore, provides useful information about the disease process. It is a tool designed primarily for group level studies, however. Although it has been applied to assess patterns of atrophy in single patients, it has not yet been optimized fully for single-patient comparisons. Once these optimizations have occurred, VBM could be a powerful tool in the differential diagnosis of individual patients.

Hippocampal atrophy also is shown to occur several years before patients are diagnosed with AD [5]. Patterns similar to those observed in AD also are present in patients who have mild cognitive impairment (MCI). This syndrome is considered a transitional period between normal ageing and a diagnosis of AD, with patients showing early deficits in memory but not fulfilling criteria for dementia. These patients show atrophy of the medial temporal lobe structures, usually at a level intermediate between those of controls and AD (see Fig. 1). Patients who have MCI and are at greater genetic risk have smaller hippocampal volumes [6], and atrophy of hippocampus and amygdala can predict progression to dementia in cognitively intact elderly individuals and in patients who have MCI [7], although the overlap is too large to have a prognostic value in individual patients.

Structural MRI also can help in the differential diagnosis of AD from other neurodegenerative dementias, especially frontotemporal lobar degeneration (FTLD). There is some overlap in the clinical features of the two dementias, which increases the importance of imaging in the clinical assessment. Patients who have FTLD show more severe atrophy in the frontal lobe and anterior temporal poles than patients who have AD, with little involvement of the temporoparietal association neocortex. This anterior-posterior gradient of atrophy in the brains of patients who have FTLD can help distinguish them from subjects with AD, who show a more posterior bias [8]. Asymmetry also is a common feature that is relatively specific to FTLD and can be useful diagnostically. FTLD can be divided into several different syndromic variants that show different, although overlapping, patterns of atrophy [9], [10]. The behavioral variant of frontotemporal dementia (bvFTD) presents with early behavioral abnormalities and is associated with atrophy of the frontal and temporal lobes, although the frontal lobes often show the greatest loss. In contrast, the other two variants present with early language deficits. Patients who have semantic dementia present with a loss of memory for words and show a well-defined pattern of atrophy affecting the anterior temporal lobes most predominantly (Fig. 3) [11]. The patterns often are asymmetric, affecting the left temporal lobe in particular, although the right temporal lobes also can show the brunt of the loss. The patterns observed in progressive nonfluent aphasia, the other language variant, are more variable but generally involve almost exclusively the left hemisphere, in particular the regions of the brain surrounding the perisylvian fissure, including the inferior frontal lobe and insula [10].

The patterns of atrophy present in patients who have dementia with Lewy bodies (DLB) are not as well established and visual inspection of individual magnetic resonance scans for a specific atrophy pattern typically is not used to aid clinical diagnosis of DLB. Some group studies have shown patterns of atrophy similar to those observed in AD, although others have observed a more focal pattern of loss in the basal forebrain in DLB. Patients who have DLB do seem to show less atrophy of the medial temporal lobe, which may prove diagnostically useful [12]. Imaging, however, is necessary for the diagnosis of vascular dementia (VaD) [12]. MRI, in particular, is more sensitive to vascular changes than CT, especially in subcortical regions. The features characteristic of VaD include cortical infarctions, lacunar infarctions, and diffuse white matter hyperintensities, also known as leukariosis, that appear bright on a T2-weighted or fluid-attenuated inversion recovery (FLAIR) sequence and reflect regions of demyelination and enlargement of perivascular spaces (Fig. 4). Although the presence of lacunar infarctions and white matter hyperintensities defines VaD, they also often are present, although to a lesser degree, in healthy elderly controls and in patients who have AD and FTLD. The presence of high signal abnormalities themselves, therefore, is not particularly diagnostically useful, and can correctly classify only 42% of patients who have VaD on visual inspection [13]. Cerebral atrophy does occur in patients who have VaD, although not to the same degree as brain losses in AD. No specific patterns of atrophy have been observed, with mild losses reported in the frontal, lateral temporal, medial temporal, and parietal lobes [14]. Characteristic imaging features also can be identified in patients who have Creutzfeldt-Jakob disease. Increased signal intensity often is observed bilaterally in the basal ganglia and cortical ribbon on a FLAIR sequence (Fig. 5).

There is increasing interest in studying change in the brain over multiple serial MRI scans in patients who have dementia. Rates of atrophy can be calculated from pairs of scans that have been matched positionally or registered. There is some evidence that rates of cerebral atrophy may aid in the differentiation of different neurodegenerative dementias [15]. Rates of cerebral atrophy in AD are reported in the range of 1.5% to 3% per year, with rates of hippocampal atrophy approximately 4% to 6% per year. Rates of hippocampal atrophy can differentiate AD from controls with a high sensitivity and specificity and have a greater discriminatory power than baseline volumes [16]. There also is evidence that high rates of atrophy in patients who have MCI help predict conversion to AD [17]. Although these techniques currently are not used in clinical practice, the procedures are relatively simple and could be applied if the technique was made easy to use and automated. Rates of whole brain and regional atrophy increasingly are incorporated into clinical drug trials as one of the primary outcome measures. More complex registration methods also have been developed that can provide detailed information concerning exactly where in the brain change has occurred over time. These techniques show widespread patterns of tissue loss in the temporal and parietal lobes, with a relative sparing of the sensorimotor cortices, over time in patients who have AD [18], [19]. These techniques are not widely available, however they are automated, would be relatively easy to apply in a clinical setting, and would provide invaluable diagnostic information about the progression of brain atrophy in individual patients.

• Functional nuclear medicine

In contrast to structural imaging, functional nuclear medicine techniques measure glucose uptake or cerebral blood flow. It is hypothesized that functional losses may precede structural changes in the brain and, therefore, may provide more sensitive markers of early disease progression. There are two main techniques that this article considers. F18-fluorodeoxyglucose–positron emission tomography (FDG-PET) measures the local cerebral metabolic rate of glucose uptake. In contrast, single photon emission CT (SPECT) measures blood flow alterations, or perfusion. Both techniques require the injection of radioactive tracers. Imaging using PET has a higher sensitivity and greater spatial resolution than SPECT. The Centers for Medicare and Medicaid Services have approved FDG-PET imaging as a routine examination tool for the early and differential diagnosis of AD specifically to differentiate AD from FTLD. This was based on evidence showing that adding PET to a clinical examination increases diagnostic sensitivity in AD.

It is well established that by the time a patient presents with clinical dementia symptoms, a reduction in glucose uptake already has occurred. FDG-PET shows a pattern of abnormally low uptake in the posterior cingulate, precuneus, temporoparietal regions, and frontal cortex in AD when compared with controls (Fig. 6). Hypometabolism also can be observed in the medial temporal lobe [20]. These patterns can differentiate patients who have pathologically confirmed AD from controls with sensitivities and specificities of approximately 86%, although they are as high as 100% in some studies [21]. Similar patterns of hypoperfusion have been identified on SPECT in patients who have AD, although it has been shown that PET provides better differentiation between AD and control patients.

FDG-PET also can detect very early changes in uptake. Reports of patients at increased risk for AD, either because of family history or genetic susceptibility, show reductions in glucose uptake before the onset of clinical symptoms [22]. Studies in MCI show medial temporal and temporoparietal hypometabolism on FDG-PET (see Fig. 6). Visual assessments of the medial temporal lobe can distinguish patients who have MCI from controls at a sensitivity of 77% and a specificity of 71% [20]. Temporoparietal hypometabolism in MCI also is shown to predict conversion to AD with an accuracy of 75% to 100% [21]. Hypoperfusion has been demonstrated in patients who have MCI on SPECT, particularly in the posterior cingulate, although only minor changes can be observed on visual inspection.

A significant value of PET is expected to be in the differential diagnosis of AD and other neurodegenerative conditions. The patterns of metabolic and perfusion abnormalities in AD are different from those observed in patients who have FTD, which show early and more severe frontal and anterior/mesial temporal hypometabolism and increased perfusion [23], [24]. These patterns often are asymmetric in patients who have FTLD [14]. The signal of decreased perfusion in the posterior cingulate that is useful in the diagnosis of AD is not present in FTLD on SPECT [25]. There also is some evidence that the different syndromic variants of FTLD show different patterns of hypometabolism and hypoperfusion. There are some trends for left inferior frontal changes in progressive nonfluent aphasia [26], left anterior temporal changes in semantic dementia, and frontal lobe changes in bvFTD [27], although there is considerable overlap between the different variants and the value of these trends in differential diagnosis is unknown. The patterns of hypometabolism and hypoperfusion observed in patients who have DLB closely mirror those observed in AD, although patients who have DLB seem to have greater reductions in the occipital lobe, particularly in the primary visual cortex, compared with AD [28], [29]. Patterns of brain metabolic activity are less well defined for patients who have VaD. Hypometabolism has been observed in cortical regions but also in subcortical regions and the cerebellum, which usually are spared in AD [30], although others have found no characteristic patterns in patients who have VaD.

• Functional MRI

Functional MRI (fMRI) uses the paramagnetic properties of oxygenated versus deoxygenated blood to visualize areas of cortex engaged in specific activation tasks. Neural activation causes a proportionately greater increase in regional blood flow than in oxygen consumption. This produces an increase in the local cerebral blood oxygenation level and a decrease in deoxyhemoglobin. The magnetic resonance signal of blood is slightly different depending on the level of oxygenation. The magnitude of the signal difference, however, is too small to be identified on single images; therefore, multiple scans are performed and summed. This technique has higher spatial resolution and is less expensive and invasive than PET; however, at the moment it remains largely a research tool.

Studies show reduced activation in regions, including the medial temporal lobes, during memory tasks in patients who have MCI or AD compared with groups of control subjects [31], reflecting neuronal loss or dysfunction in these regions [32], [33]. Differences in activity pattern also are observed between patients who are at risk for AD and those who are not [34], [35]. There are suggestions that patients who have FTLD show reduced activation in the frontal lobes and patients who have DLB show reduced activation in the occipital lobes during a variety of cognitive tasks compared with AD. Although most fMRI studies aim to identify areas of increased activation related to specific tasks, several recent studies have identified a network of regions that are more active during periods of mental rest from specific cognitive tasks (ie, the “default network”). Default activation patterns in young adults correspond to regions that show amyloid deposition, atrophy, or hypometabolism on PET [36]. Patients who have AD or MCI show decreased activation in the default network, in regions including the hippocampus, posterior cingulate, and precuneus [37].

• 1H magnetic resonance spectroscopy

1H magnetic resonance spectroscopy (MRS) is a noninvasive technique that allows assessment of specific brain metabolites. It is unique among diagnostic imaging techniques in that it allows the signals from several different metabolites to be measured within a single measurement period, with each metabolite in turn sensitive to a different aspect of in vivo pathologic processes at the molecular or cellular level. The metabolites measured most commonly include N-acetylaspartate (NAA), which provides a marker of neuronal density; myoinositol (mI), which provides a marker of glial cell activity; and choline (Cho), which is believed to reflect the level of membrane turnover.

It is well established that patients who have AD show a decrease in the level of NAA in several brain regions, including the posterior cingulate, temporal, and parietal and frontal lobes, compared with control subjects. In contrast, the levels of mI increase in patients who have AD. The clinical specificity of the NAA decline is poor, but a decrease in the ratio of the NAA to mI is robust in discriminating patients who have AD from healthy controls. The metabolite changes in patients who have MCI generally are intermediate between normal elderly controls and patients who have AD [38] and can predict the rate at which those patients will progress to AD [39]. Longitudinal studies also show progressive decreases in NAA over time, which seem to correlate to clinical decline in AD [40]. Differences in the profile and regional distribution of metabolites are observed in other dementias, which may aid in the differential diagnosis [41]. Patients who have FTLD show decreased levels of NAA and Cho and increased levels of mI in the frontal lobes, and decreased levels of NAA in the temporal lobes, compared with subjects who have AD [42], [43]. In contrast to AD and FTLD, the levels of NAA are not decreased in the gray matter of patients who have DLB; the only difference from controls is in the Cho levels that are elevated [41], probably reflecting the fact that severe cholinergic deficits are a feature of DLB. Patients who have VaD show decreased levels of NAA, as in AD and FTLD, although the mI and Cho levels usually are not elevated [41]. MRS, therefore, could contribute to the differential diagnosis and early detection of AD and monitoring disease progression. Results from serial MRS studies, however, are mixed and need further validation before MRS can be considered a biomarker for disease progression.

• Diffusion-weighted imaging

The technique of diffusion-weighted imaging (DWI) allows the measurement of the microscopic random motion of water molecules in the brain. The more modern version of DWI is diffusion tensor imaging. DWI and diffusion tensor imaging produce measures of a variety of features of water diffusion but the two studied most widely are apparent diffusion coefficient (ADC) and fractional anisotropy. Both are scalar parameters. ADC is a measure of diffusion magnitude, whereas fractional anisotropy is an indication of the directionality of water diffusion. Increases occur in ADC and decreases in fractional anisotropy in degenerative brains as a result of an assumed loss or disruption of barriers restricting water motion, such as the membranes of cell bodies, axons, and myelin. Patients who have AD and MCI show elevated ADC in brain regions that typically are involved in AD. In addition, hippocampal diffusivity is greater in patients who have MCI and who convert to dementia compared with those who remain stable and may provide better prediction of rate of conversion than hippocampal volumes measured on MRI [44]. Diffusion-weighted imaging has potential, therefore, to be clinically useful but as with other imaging techniques, it remains purely a research tool at the present time.

• Amyloid imaging

A recent major advance in the field of neuroimaging is the development of a technique to image amyloid in the human brain [45]. The presence of amyloid plaques and neurofibrillary tangles are the hallmark pathologic features of AD and are required to give a definite pathologic diagnosis of AD. The ability to identify plaques in living patients has the potential to revolutionize patient diagnosis and management.

Several amyloid-binding compounds have been developed and studied in humans [46], [47], [48]. The compound studied the most extensively is Pittsburgh Compound-B (PIB). PIB binds to aggregated fibrillar amyloid-beta deposits with high affinity and can be detected by PET imaging [47]. A robust signal has been identified in patients who have AD that is different from that seen in healthy controls (Fig. 7). In patients who have AD, PIB binding is most prominent in cortical association areas, including the frontal lobes, temporal lobes, parietal lobes, parts of occipital lobe, and striatum [47], [49]. Amyloid deposition is found commonly on pathology in all these regions in patients who have AD. Particularly severe binding has been observed in the frontal lobes [47] and precuneus [49], with low levels observed in the sensorimotor strip, primary visual cortex, and medial temporal lobe [49]. An absence of binding has been observed in the cerebellum, pons, and subcortical white matter, areas that do not show amyloid deposition on pathology. These changes are present in practically but not all patients who have clinically evident AD. It is possible that the few patients who do not show PIB retention may have a different underlying cause for their dementia. Similar patterns of retention have been observed in sporadic and familial patients who have AD, although there is some evidence that neostriatal PIB binding is higher in familial patients. The levels of PIB retention, however, seem stable over time, even in patients who have declined cognitively [50].

In comparison, the majority of healthy controls show little or no retention in cortical areas [47]. Some studies, however, have reported healthy controls who do show positive PIB binding, sometimes at a similar level to that found in AD. This is consistent with the fact that neuropathologic AD is present in some nondemented elderly subjects and may suggest that the pathologic changes of AD can be detected before any clinical evidence of dementia. In support of this hypothesis is neocortical PIB binding that has been observed in presymptomatic subjects who have familial AD. PIB retention seems to vary greatly in MCI, with patients often showing a pattern of binding similar to AD or similar to controls. Patients rarely show intermediate levels of PIB binding (see Fig. 7) [51].

The assessment of PIB imaging also may prove highly useful in the differential diagnosis of dementia, especially in differentiating AD from FTLD. Recent work suggests that no PIB binding is found in patients who have FTLD [52]. A pattern of binding similar to that found in AD is observed in DLB, although it is present in only 89% of patients [52]. This most likely reflects the fact that the majority of DLB patients have concomitant AD pathology.

Therefore, although research in amyloid imaging in humans is in its infancy, it shows huge promise as a sensitive and specific marker of AD amyloid pathology. It has the potential to revolutionize clinical practice and the conduct of clinical trials completely, as a measure of amyloid reduction and as a potential inclusion criterion. It is still important, however, to be cautious, because large studies with long clinical follow-up, or extensive postmortem confirmation, have not yet been performed. This validation will be crucial before PIB can be integrated into clinical practice.

SUMMARY

Over the past decade there has been an exponential increase in the number of studies applying neuroimaging techniques to the study of degenerative dementia. A diverse range of techniques is available, some of which already are used routinely in clinical practice, whereas others still are research tools but hold significant promise for the future. Structural images already are an important component of the clinical assessment, as they allow the identification and exclusion of potentially treatable causes of cognitive impairment. They also highlight structural changes in the brain, which can be useful diagnostically. Many other techniques also have been developed to measure aspects of brain function, such as rates of glucose uptake or cerebral blood flow and the levels of certain brain metabolites. Imaging amyloid in the brain of living patients also is possible now and is likely to revolutionize patient diagnosis and management. A huge amount of imaging work has focused on patients who have AD, although increasingly research also is beginning to focus on the non-AD dementias, such as FTLD, DLB, and VaD. There are several key aspects of clinical practice for which imaging can play an important role: first, in the differential diagnosis of the different dementias; second, in allowing early detection and prediction of patients who will develop dementia; and third, in monitoring progression of the disease over time. In addition to clinical usefulness, an important aim is to develop surrogate markers of disease progression that can be used in the assessment of potentially disease-modifying therapies.

---

References

[1] [1]Wahlund LO, Julin P, Johansson SE, et al.. Visual rating and volumetry of the medial temporal lobe on magnetic resonance imaging in dementia: a comparative study. J Neurol Neurosurg Psychiatry. 2000;69(5):630–635.

[2] [2]Jack CR, Dickson DW, Parisi JE, et al.. Antemortem MRI findings correlate with hippocampal neuropathology in typical aging and dementia. Neurology. 2002;58(5):750–757.

[3] [3]Whitwell JL, Jack CR. Comparisons between Alzheimer’s disease, frontotemporal lobar degeneration, and normal ageing with brain mapping. Top Magn Reson Imaging. 2006;16:409–425.

[4] [4]Frisoni GB, Testa C, Sabattoli F, et al.. Structural correlates of early and late onset Alzheimer’s disease: voxel based morphometric study. J Neurol Neurosurg Psychiatry. 2005;76(1):112–114.

[5] [5]Fox NC, Warrington EK, Freeborough PA, et al.. Presymptomatic hippocampal atrophy in Alzheimer’s disease. A longitudinal MRI study. Brain. 1996;119(Pt 6):2001–2007.

[6] [6]Farlow MR, He Y, Tekin S, et al.. Impact of APOE in mild cognitive impairment. Neurology. 2004;63(10):1898–1901.

[7] [7]Jack CR, Petersen RC, Xu YC, et al.. Prediction of AD with MRI-based hippocampal volume in mild cognitive impairment. Neurology. 1999;52(7):1397–1403.

[8] [8]Likeman M, Anderson VM, Stevens JM, et al.. Visual assessment of atrophy on magnetic resonance imaging in the diagnosis of pathologically confirmed young-onset dementias. Arch Neurol. 2005;62(9):1410–1415.

[9] [9]Rosen HJ, Gorno-Tempini ML, Goldman WP, et al.. Patterns of brain atrophy in frontotemporal dementia and semantic dementia. Neurology. 2002;58(2):198–208.

[10] [10]Josephs KA, Duffy JR, Strand EA, et al.. Clinicopathological and imaging correlates of progressive aphasia and apraxia of speech. Brain. 2006;129:1385–1398.

[11] [11]Chan D, Fox NC, Scahill RI, et al.. Patterns of temporal lobe atrophy in semantic dementia and Alzheimer’s disease. Ann Neurol. 2001;49(4):433–442.

[12] [12]Whitwell JL, Weigand SD, Shiung MM, et al.. Focal atrophy in dementia with Lewy bodies on MRI: a distinct pattern from Alzheimer’s disease. Brain. 2007;130:708–719.

[13] [13]Varma AR, Laitt R, Lloyd JJ, et al.. Diagnostic value of high signal abnormalities on T2 weighted MRI in the differentiation of Alzheimer’s, frontotemporal and vascular dementias. Acta Neurol Scand. 2002;105(5):355–364.

[14] [14]Varma AR, Adams W, Lloyd JJ, et al.. Diagnostic patterns of regional atrophy on MRI and regional cerebral blood flow change on SPECT in young onset patients with Alzheimer’s disease, frontotemporal dementia and vascular dementia. Acta Neurol Scand. 2002;105(4):261–269.

[15] [15]Whitwell JL, Jack CR, et al.. Rates of cerebral atrophy differ in different degenerative pathologies. Brain. 2007;130:1148–1158.

[16] [16]Barnes J, Scahill RI, Boyes RG, et al.. Differentiating AD from aging using semiautomated measurement of hippocampal atrophy rates. Neuroimage. 2004;23(2):574–581.

[17] [17]Jack CR, Shiung MM, Weigand SD, et al.. Brain atrophy rates predict subsequent clinical conversion in normal elderly and amnestic MCI. Neurology. 2005;65(8):1227–1231.

[18] [18]Fox NC, Crum WR, Scahill RI, et al.. Imaging of onset and progression of Alzheimer’s disease with voxel-compression mapping of serial magnetic resonance images. Lancet. 2001;358(9277):201–205.

[19] [19]Thompson PM, Hayashi KM, de Zubicaray G, et al.. Dynamics of gray matter loss in Alzheimer’s disease. J Neurosci. 2003;23(3):994–1005.

[20] [20]Mosconi L, De Santi S, Li Y, et al.. Visual rating of medial temporal lobe metabolism in mild cognitive impairment and Alzheimer’s disease using FDG-PET. Eur J Nucl Med Mol Imaging. 2006;33(2):210–221.

[21] [21]Mosconi L. Brain glucose metabolism in the early and specific diagnosis of Alzheimer’s disease. FDG-PET studies in MCI and AD. Eur J Nucl Med Mol Imaging. 2005;32(4):486–510.

[22] [22]Jagust W, Gitcho A, Sun F, et al.. Brain imaging evidence of preclinical Alzheimer’s disease in normal aging. Ann Neurol. 2006;59(4):673–681.

[23] [23]Ibach B, Poljansky S, Marienhagen J, et al.. Contrasting metabolic impairment in frontotemporal degeneration and early onset Alzheimer’s disease. Neuroimage. 2004;23(2):739–743.

[24] [24]Jeong Y, Cho SS, Park JM, et al.. 18F-FDG PET findings in frontotemporal dementia: an SPM analysis of 29 patients. J Nucl Med. 2005;46(2):233–239.

[25] [25]Bonte FJ, Harris TS, Roney CA, et al.. Differential diagnosis between Alzheimer’s and frontotemporal disease by the posterior cingulate sign. J Nucl Med. 2004;45(5):771–774.

[26] [26]Zahn R, Buechert M, Overmans J, et al.. Mapping of temporal and parietal cortex in progressive nonfluent aphasia and Alzheimer’s disease using chemical shift imaging, voxel-based morphometry and positron emission tomography. Psychiatry Res. 2005;140(2):115–131.

[27] [27]McMurtray AM, Chen AK, Shapira JS, et al.. Variations in regional SPECT hypoperfusion and clinical features in frontotemporal dementia. Neurology. 2006;66(4):517–522.

[28] [28]Minoshima S, Foster NL, Sima AA, et al.. Alzheimer’s disease versus dementia with Lewy bodies: cerebral metabolic distinction with autopsy confirmation. Ann Neurol. 2001;50(3):358–365.

[29] [29]Gilman S, Koeppe RA, Little R, et al.. Differentiation of Alzheimer’s disease from dementia with Lewy bodies utilizing positron emission tomography with [18F]fluorodeoxyglucose and neuropsychological testing. Exp Neurol. 2005;191(Suppl 1):S95–S103.

[30] [30]Kerrouche N, Herholz K, Mielke R, et al.. (18)FDG PET in vascular dementia: differentiation from Alzheimer’s disease using voxel-based multivariate analysis. J Cereb Blood Flow Metab. 2006;26(9):1213–1221.

[31] [31]Sperling RA, Bates JF, Chua EF, et al.. fMRI studies of associative encoding in young and elderly controls and mild Alzheimer’s disease. J Neurol Neurosurg Psychiatry. 2003;74(1):44–50.

[32] [32]Machulda MM, Ward HA, Borowski B, et al.. Comparison of memory fMRI response among normal, MCI, and Alzheimer’s patients. Neurology. 2003;61(4):500–506.

[33] [33]Johnson SC, Schmitz TW, Moritz CH, et al.. Activation of brain regions vulnerable to Alzheimer’s disease: the effect of mild cognitive impairment. Neurobiol Aging. 2005;27(11):1604–1612.

[34] [34]Bookheimer SY, Strojwas MH, Cohen MS, et al.. Patterns of brain activation in people at risk for Alzheimer’s disease. N Engl J Med. 2000;343(7):450–456.

[35] [35]Johnson SC, Schmitz TW, Trivedi MA, et al.. The influence of Alzheimer disease family history and apolipoprotein E epsilon4 on mesial temporal lobe activation. J Neurosci. 2006;26(22):6069–6076.

[36] [36]Buckner RL, Snyder AZ, Shannon BJ, et al.. Molecular, structural, and functional characterization of Alzheimer’s disease: evidence for a relationship between default activity, amyloid, and memory. J Neurosci. 2005;25(34):7709–7717.

[37] [37]Rombouts SA, Barkhof F, Goekoop R, et al.. Altered resting state networks in mild cognitive impairment and mild Alzheimer’s disease: an fMRI study. Hum Brain Mapp. 2005;26(4):231–239.

[38] [38]Kantarci K, Jack CR, Xu YC, et al.. Regional metabolic patterns in mild cognitive impairment and Alzheimer’s disease: a 1H MRS study. Neurology. 2000;55(2):210–217.

[39] [39]Modrego PJ, Fayed N, Pina MA. Conversion from mild cognitive impairment to probable Alzheimer’s disease predicted by brain magnetic resonance spectroscopy. Am J Psychiatry. 2005;162(4):667–675.

[40] [40]Kantarci K, Weigand SD, Petersen RC, et al.. Longitudinal (1)H MRS changes in mild cognitive impairment and Alzheimer’s disease. [Epub] Neurobiol Aging. 2006;129(11):2856–2866.

[41] [41]Kantarci K, Petersen RC, Boeve BF, et al.. 1H MR spectroscopy in common dementias. Neurology. 2004;63(8):1393–1398.

[42] [42]Mihara M, Hattori N, Abe K, et al.. Magnetic resonance spectroscopic study of Alzheimer’s disease and frontotemporal dementia/Pick complex. Neuroreport. 2006;17(4):413–416.

[43] [43]Coulthard E, Firbank M, English P, et al.. Proton magnetic resonance spectroscopy in frontotemporal dementia. J Neurol. 2006;253(7):861–868.

[44] [44]Kantarci K, Petersen RC, Boeve BF, et al.. DWI predicts future progression to Alzheimer disease in amnestic mild cognitive impairment. Neurology. 2005;64(5):902–904.

[45] [45]Nordberg A. PET imaging of amyloid in Alzheimer’s disease. Lancet Neurol. 2004;3(9):519–527.

[46] [46]Shoghi-Jadid K, Small GW, Agdeppa ED, et al.. Localization of neurofibrillary tangles and beta-amyloid plaques in the brains of living patients with Alzheimer disease. Am J Geriatr Psychiatry. 2002;10(1):24–35.

[47] [47]Klunk WE, Engler H, Nordberg A, et al.. Imaging brain amyloid in Alzheimer’s disease with Pittsburgh Compound-B. Ann Neurol. 2004;55(3):306–319.

[48] [48]Newberg AB, Wintering NA, Plossl K, et al.. Safety, biodistribution, and dosimetry of 123I-IMPY: a novel amyloid plaque-imaging agent for the diagnosis of Alzheimer’s disease. J Nucl Med. 2006;47(5):748–754.

[49] [49]Mintun MA, Larossa GN, Sheline YI, et al.. [11C] PIB in a nondemented population: potential antecedent marker of Alzheimer disease. Neurology. 2006;67(3):446–452.

[50] [50]Engler H, Forsberg A, Almkvist O, et al.. Two-year follow-up of amyloid deposition in patients with Alzheimer’s disease. Brain. Jul 19 2006;.

[51] [51]Lopresti BJ, Klunk WE, Mathis CA, et al.. Simplified quantification of Pittsburgh compound B amyloid imaging PET studies: a comparative analysis. J Nucl Med. 2005;46(12):1959–1972.

[52] [52]Rowe CR, Ng S, Gong SJ, et al.. C-11 PIB PET amyloid imaging in aging and dementia. Alzheimer’s & Dementia. 2006;2(3):S66.