Atrophy of the medial occipitotemporal, inferior, and middle temporal gyri in non-demented elderly predict decline to Alzheimer's disease

Regional manifold learning for disease classification

While manifold learning from images itself has become widely used in medical image analysis, the accuracy of existing implementations suffers from viewing each image as a single data point. To address this issue, we parcellate images into regions and then separately learn the manifold for each region. We use the regional manifolds as low-dimensional descriptors of high-dimensional morphological image features, which are then fed into a classifier to identify regions affected by disease. We produce a single ensemble decision for each scan by the weighted combination of these regional classification results. Each weight is determined by the regional accuracy of detecting the disease. When applied to cardiac magnetic resonance imaging of 50 normal controls and 50 patients with reconstructive surgery of Tetralogy of Fallot, our method achieves significantly better classification accuracy than approaches learning a single manifold across the entire image domain.

Latent information in fluency lists predicts functional decline in persons at risk for Alzheimer disease

OBJECTIVE

We constructed random forest classifiers employing either the traditional method of scoring semantic fluency word lists or new methods. These classifiers were then compared in terms of their ability to diagnose Alzheimer disease (AD) or to prognosticate among individuals along the continuum from cognitively normal (CN) through mild cognitive impairment (MCI) to AD.

METHOD

Semantic fluency lists from 44 cognitively normal elderly individuals, 80 MCI patients, and 41 AD patients were transcribed into electronic text files and scored by four methods: traditional raw scores, clustering and switching scores, "generalized" versions of clustering and switching, and a method based on independent components analysis (ICA). Random forest classifiers based on raw scores were compared to "augmented" classifiers that incorporated newer scoring methods. Outcome variables included AD diagnosis at baseline, MCI conversion, increase in Clinical Dementia Rating-Sum of Boxes (CDR-SOB) score, or decrease in Financial Capacity Instrument (FCI) score. Receiver operating characteristic (ROC) curves were constructed for each classifier and the area under the curve (AUC) was calculated. We compared AUC between raw and augmented classifiers using Delong's test and assessed validity and reliability of the augmented classifier.

RESULTS

Augmented classifiers outperformed classifiers based on raw scores for the outcome measures AD diagnosis (AUC .97 vs. .95), MCI conversion (AUC .91 vs. .77), CDR-SOB increase (AUC .90 vs. .79), and FCI decrease (AUC .89 vs. .72). Measures of validity and stability over time support the use of the method.

CONCLUSION

Latent information in semantic fluency word lists is useful for predicting cognitive and functional decline among elderly individuals at increased risk for developing AD. Modern machine learning methods may incorporate latent information to enhance the diagnostic value of semantic fluency raw scores. These methods could yield information valuable for patient care and clinical trial design with a relatively small investment of time and money.

Matrix-Similarity Based Loss Function and Feature Selection for Alzheimer's Disease Diagnosis

Recent studies on Alzheimer's Disease (AD) or its prodromal stage, Mild Cognitive Impairment (MCI), diagnosis presented that the tasks of identifying brain disease status and predicting clinical scores based on neuroimaging features were highly related to each other. However, these tasks were often conducted independently in the previous studies. Regarding the feature selection, to our best knowledge, most of the previous work considered a loss function defined as an element-wise difference between the target values and the predicted ones. In this paper, we consider the problems of joint regression and classification for AD/MCI diagnosis and propose a novel matrix-similarity based loss function that uses high-level information inherent in the target response matrix and imposes the information to be preserved in the predicted response matrix. The newly devised loss function is combined with a group lasso method for joint feature selection across tasks, i.e., clinical scores prediction and disease status identification. We conducted experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, and showed that the newly devised loss function was effective to enhance the performances of both clinical score prediction and disease status identification, outperforming the state-of-the-art methods.

Tract-based spatial statistics: application to mild cognitive impairment

RATIONALE AND OBJECTIVES

The primary objective of the current investigation was to characterize white matter integrity in different subtypes of mild cognitive impairment (MCI) using tract-based spatial statistics of diffusion tensor imaging.

MATERIALS AND METHODS

The study participants were divided into 4 groups of 30 subjects each as follows: cognitively healthy controls, amnestic MCI, dysexecutive MCI, and Alzheimer's disease (AD). All subjects underwent a comprehensive neuropsychological assessment, apolipoprotein E genotyping, and 3-tesla MRI. The diffusion tensor was reconstructed and then analyzed using tract-based spatial statistics. The changes in brain white matter tracts were also examined according to the apolipoprotein E ε 4 status.

RESULTS

Compared with controls, amnestic MCI patients showed significant differences in the cerebral white matter, where changes were consistently detectable in the frontal and parietal lobes. We found a moderate impact of the apolipoprotein E ε 4 status on the extent of white matter disruption in the amnestic MCI group. Patients with AD exhibited similar but more extensive alterations, while no significant changes were observed in dysexecutive MCI patients.

CONCLUSION

The results from this study indicate that amnestic MCI is the most likely precursor to AD as both conditions share significant white matter damage. By contrast, dysexecutive MCI seems to be characterized by a distinct pathogenesis.

Interactive effects of apolipoprotein E4 and diabetes risk on later myelinating white matter regions in neurologically healthy older aged adults

Possession of the apolipoprotein E4 (APOE4) allele and diabetes risk are independently related to reduced white matter (WM) integrity that may contribute to the development of Alzheimer's disease (AD). The purpose of this study is to examine the interactive effects of APOE4 and diabetes risk on later myelinating WM regions among healthy elderly individuals at risk of AD. A sample of 107 healthy elderly (80 APOE4-/27 APOE4+) individuals underwent structural magnetic resonance imaging/diffusion tensor imaging (DTI). Data were prepared using Tract-Based Spatial Statistics, and a priori regions of interest (ROIs) were extracted from T1-based WM parcellations. Regions of interest included later myelinating frontal/temporal/parietal WM regions and control regions measured by fractional anisotropy (FA). There were no APOE group differences in DTI for any ROI. Within the APOE4 group, we found negative relationships between hemoglobin A1c/fasting glucose and APOE4 on FA for all later myelinating WM regions but not for early/middle myelinating control regions. Results also showed APOE4/diabetes risk interactions for WM underlying supramarginal, superior temporal, precuneus, superior parietal, and superior frontal regions. Results suggest interactive effects of APOE4 and diabetes risk on later myelinating WM regions, which supports preclinical detection of AD among this particularly susceptible subgroup.

Neurodegenerative disease diagnosis using incomplete multi-modality data via matrix shrinkage and completion

In this work, we are interested in predicting the diagnostic statuses of potentially neurodegenerated patients using feature values derived from multi-modality neuroimaging data and biological data, which might be incomplete. Collecting the feature values into a matrix, with each row containing a feature vector of a sample, we propose a framework to predict the corresponding associated multiple target outputs (e.g., diagnosis label and clinical scores) from this feature matrix by performing matrix shrinkage following matrix completion. Specifically, we first combine the feature and target output matrices into a large matrix and then partition this large incomplete matrix into smaller submatrices, each consisting of samples with complete feature values (corresponding to a certain combination of modalities) and target outputs. Treating each target output as the outcome of a prediction task, we apply a 2-step multi-task learning algorithm to select the most discriminative features and samples in each submatrix. Features and samples that are not selected in any of the submatrices are discarded, resulting in a shrunk version of the original large matrix. The missing feature values and unknown target outputs of the shrunk matrix is then completed simultaneously. Experimental results using the ADNI dataset indicate that our proposed framework achieves higher classification accuracy at a greater speed when compared with conventional imputation-based classification methods and also yields competitive performance when compared with the state-of-the-art methods.

MTA index: a simple 2D-method for assessing atrophy of the medial temporal lobe using clinically available neuroimaging

Background and purpose: Despite a strong correlation to severity of AD pathology, the measurement of medial temporal lobe atrophy (MTA) is not being widely used in daily clinical practice as a criterion in the diagnosis of prodromal and probable AD. This is mainly because the methods available to date are sophisticated and difficult to implement for routine use in most hospitals—volumetric methods—or lack objectivity—visual rating scales. In this pilot study we aim to describe a new, simple and objective method for measuring the rate of MTA in relation to the global atrophy using clinically available neuroimaging and describe the rationale behind this method.

Description: This method consists of calculating a ratio with the area of 3 regions traced manually on one single coronal MRI slide at the level of the interpeduncular fossa: (1) the medial temporal lobe (MTL) region (A); (2) the parenchima within the medial temporal region, that includes the hippocampus and the parahippocampal gyrus—the fimbria taenia and plexus choroideus are excluded—(B); and (3) the body of the ipsilateral lateral ventricle (C). Therefrom we can compute the ratio “Medial Temporal Atrophy index” at both sides as follows: MTAi = (A − B)× 10/C.

Conclusions: The MTAi is a simple 2D-method for measuring the relative extent of atrophy in the MTL in relation to the global brain atrophy. This method can be useful for a more accurate diagnosis of AD in routine clinical practice. Further studies are needed to assess the usefulness of MTAi in the diagnosis of early AD, in tracking the progression of AD and in the differential diagnosis of AD with other dementias.

Group-constrained sparse fMRI connectivity modeling for mild cognitive impairment identification

Emergence of advanced network analysis techniques utilizing resting-state functional magnetic resonance imaging (R-fMRI) has enabled a more comprehensive understanding of neurological disorders at a whole-brain level. However, inferring brain connectivity from R-fMRI is a challenging task, particularly when the ultimate goal is to achieve good control-patient classification performance, owing to perplexing noise effects, curse of dimensionality, and inter-subject variability. Incorporating sparsity into connectivity modeling may be a possible solution to partially remedy this problem since most biological networks are intrinsically sparse. Nevertheless, sparsity constraint, when applied at an individual level, will inevitably cause inter-subject variability and hence degrade classification performance. To this end, we formulate the R-fMRI time series of each region of interest (ROI) as a linear representation of time series of other ROIs to infer sparse connectivity networks that are topologically identical across individuals. This formulation allows simultaneous selection of a common set of ROIs across subjects so that their linear combination is best in estimating the time series of the considered ROI. Specifically, l 1-norm is imposed on each subject to filter out spurious or insignificant connections to produce sparse networks. A group-constraint is hence imposed via multi-task learning using a l 2-norm to encourage consistent non-zero connections across subjects. This group-constraint is crucial since the network topology is identical for all subjects while still preserving individual information via different connectivity values. We validated the proposed modeling in mild cognitive impairment identification and promising results achieved demonstrate its superiority in disease characterization, particularly greater sensitivity to early stage brain pathologies. The inferred group-constrained sparse network is found to be biologically plausible and is highly associated with the disease-associated anatomical anomalies. Furthermore, our proposed approach achieved similar classification performance when finer atlas was used to parcellate the brain space.

Hierarchical manifold learning for regional image analysis

We present a novel method of hierarchical manifold learning which aims to automatically discover regional properties of image datasets. While traditional manifold learning methods have become widely used for dimensionality reduction in medical imaging, they suffer from only being able to consider whole images as single data points. We extend conventional techniques by additionally examining local variations, in order to produce spatially-varying manifold embeddings that characterize a given dataset. This involves constructing manifolds in a hierarchy of image patches of increasing granularity, while ensuring consistency between hierarchy levels. We demonstrate the utility of our method in two very different settings: 1) to learn the regional correlations in motion within a sequence of time-resolved MR images of the thoracic cavity; 2) to find discriminative regions of 3-D brain MR images associated with neurodegenerative disease.

Shape abnormalities of subcortical and ventricular structures in mild cognitive impairment and Alzheimer's disease: detecting, quantifying, and predicting

This article assesses the feasibility of using shape information to detect and quantify the subcortical and ventricular structural changes in mild cognitive impairment (MCI) and Alzheimer's disease (AD) patients. We first demonstrate structural shape abnormalities in MCI and AD as compared with healthy controls (HC). Exploring the development to AD, we then divide the MCI participants into two subgroups based on longitudinal clinical information: (1) MCI patients who remained stable; (2) MCI patients who converted to AD over time. We focus on seven structures (amygdala, hippocampus, thalamus, caudate, putamen, globus pallidus, and lateral ventricles) in 754 MR scans (210 HC, 369 MCI of which 151 converted to AD over time, and 175 AD). The hippocampus and amygdala were further subsegmented based on high field 0.8 mm isotropic 7.0T scans for finer exploration. For MCI and AD, prominent ventricular expansions were detected and we found that these patients had strongest hippocampal atrophy occurring at CA1 and strongest amygdala atrophy at the basolateral complex. Mild atrophy in basal ganglia structures was also detected in MCI and AD. Stronger atrophy in the amygdala and hippocampus, and greater expansion in ventricles was observed in MCI converters, relative to those MCI who remained stable. Furthermore, we performed principal component analysis on a linear shape space of each structure. A subsequent linear discriminant analysis on the principal component values of hippocampus, amygdala, and ventricle leads to correct classification of 88% HC subjects and 86% AD subjects.

Time course of brain volume changes in the preclinical phase of Alzheimer's disease

BACKGROUND

Structural alterations of a large network characterize Alzheimer's disease (AD), but the time course of these changes remains unclear. The dynamic of these alterations was examined in the AD preclinical phase using data from the 10-year follow-up of a population-based cohort (Bordeaux-3City).

METHODS

Participants received neuropsychological assessments every 2 years and two identical magnetic resonance imaging (MRI) exams at baseline and 4 years later. Twenty-five incident AD cases were compared with 319 subjects who remained free of dementia. Subjects were free of dementia at baseline and at follow-up MRI. Incident AD occurred after these time points.

RESULTS

At baseline, incident AD already presented smaller volumes only in the left amygdalo-hippocampal complex. Moreover, a higher annual rate of atrophy of the temporoparietal cortices was observed in future AD subjects during the following 4 years.

CONCLUSION

Incident AD cases present mediotemporal lesions up to 5 years before diagnosis. This neurodegenerative process seems to progressively reach the temporoparietal cortices in the AD preclinical phase.

Connectome-scale assessments of structural and functional connectivity in MCI

Mild cognitive impairment (MCI) has received increasing attention not only because of its potential as a precursor for Alzheimer's disease but also as a predictor of conversion to other neurodegenerative diseases. Although MCI has been defined clinically, accurate and efficient diagnosis is still challenging. Although neuroimaging techniques hold promise, compared to commonly used biomarkers including amyloid plaques, tau protein levels and brain tissue atrophy, neuroimaging biomarkers are less well validated. In this article, we propose a connectomes-scale assessment of structural and functional connectivity in MCI via two independent multimodal DTI/fMRI datasets. We first used DTI-derived structural profiles to explore and tailor the most common and consistent landmarks, then applied them in a whole-brain functional connectivity analysis. The next step fused the results from two independent datasets together and resulted in a set of functional connectomes with the most differentiation power, hence named as "connectome signatures." Our results indicate that these "connectome signatures" have significantly high MCI-vs-controls classification accuracy, at more than 95%. Interestingly, through functional meta-analysis, we found that the majority of "connectome signatures" are mainly derived from the interactions among different functional networks, for example, cognition-perception and cognition-action domains, rather than from within a single network. Our work provides support for using functional "connectome signatures" as neuroimaging biomarkers of MCI.

Mild cognitive impairment

PURPOSE OF REVIEW

The term mild cognitive impairment (MCI) is used to describe older subjects with demonstrable cognitive impairment who have not crossed the threshold for dementia. Because patients with MCI have an increased risk of developing dementia, especially Alzheimer disease (AD), there is significant interest in the clinical characterization of these subjects and in understanding the pathophysiology of the transition from MCI to AD.

RECENT FINDINGS

The MCI syndrome, as an expression of an incipient disorder that may lead to dementia, is extremely heterogeneous and may coexist with systemic, neurologic, or psychiatric disorders that can cause cognitive deficits. Recent clinical criteria were designed to take into account the different forms of clinical presentation of the syndrome, and introduced the possible contribution of biomarkers to the clinical diagnosis. Bedside diagnosis of MCI can be difficult, since patients who report having cognitive problems may have normal scores in global cognitive scales or in brief neuropsychological instruments.

SUMMARY

This article presents the evolution of the clinical concept of MCI, the operationalization of its current definitions, the development of biomarkers that can help to identify an underlying neurodegenerative process as the etiology of the syndrome, and its proposed treatments.

A combined electrophysiological and morphological examination of episodic memory decline in amnestic mild cognitive impairment

Early stages of Alzheimer’s disease (AD) are characterized by neuropathological changes within the medial temporal lobe cortex (MTLC), which lead to characteristic impairments in episodic memory, i.e., amnestic mild cognitive impairment (aMCI). Here, we tested the neural correlates of this memory impairment using event-related potentials (ERPs) and voxel-based morphometry. Twenty-four participants were instructed to encode lists of words and were tested in a yes/no recognition memory task. The dual-process model of recognition memory dissociates between acontextual familiarity and recollection of contextual details. The early frontal ERP old/new effect, which is thought to represent a neural correlate of familiarity-based memory, was absent in aMCI, whereas the control group showed a significant early old/new effect at frontal electrodes. This effect was positively correlated with behavioral episodic memory performance. Analyses of brain morphology revealed a focused gray matter loss in the inferior and medial temporal lobes in aMCI versus healthy controls. Moreover, the positive correlation between gray matter volume in the MTLC and the familiarity-related early frontal old/new effect supports the notion that this effect relies upon the integrity of the MTLC. Thus, the present findings might provide a further functional marker for prodromal AD.

Evaluating Alzheimer's disease progression using rate of regional hippocampal atrophy

Alzheimer's disease (AD) is characterized by neurofibrillary tangle and neuropil thread deposition, which ultimately results in neuronal loss. A large number of magnetic resonance imaging studies have reported a smaller hippocampus in AD patients as compared to healthy elderlies. Even though this difference is often interpreted as atrophy, it is only an indirect measurement. A more direct way of measuring the atrophy is to use repeated MRIs within the same individual. Even though several groups have used this appropriate approach, the pattern of hippocampal atrophy still remains unclear and difficult to relate to underlying pathophysiology. Here, in this longitudinal study, we aimed to map hippocampal atrophy rates in patients with AD, mild cognitive impairment (MCI) and elderly controls. Data consisted of two MRI scans for each subject. The symmetric deformation field between the first and the second MRI was computed and mapped onto the three-dimensional hippocampal surface. The pattern of atrophy rate was similar in all three groups, but the rate was significantly higher in patients with AD than in control subjects. We also found higher atrophy rates in progressive MCI patients as compared to stable MCI, particularly in the antero-lateral portion of the right hippocampus. Importantly, the regions showing the highest atrophy rate correspond to those that were described to have the highest burden of tau deposition. Our results show that local hippocampal atrophy rate is a reliable biomarker of disease stage and progression and could also be considered as a method to objectively evaluate treatment effects.

Relevance of magnetic resonance imaging for early detection and diagnosis of Alzheimer disease

Hippocampus volumetry currently is the best-established imaging biomarker for AD. However, the effect of multicenter acquisition on measurements of hippocampus volume needs to be explicitly considered when it is applied in large clinical trials, for example by using mixed-effects models to take the clustering of data within centers into account. The marker needs further validation in respect of the underlying neurobiological substrate and potential confounds such as vascular disease, inflammation, hydrocephalus, and alcoholism, and with regard to clinical outcomes such as cognition but also to demographic and socioeconomic outcomes such as mortality and institutionalization. The use of hippocampus volumetry for risk stratification of predementia study samples will further increase with the availability of automated measurement approaches. An important step in this respect will be the development of a standard hippocampus tracing protocol that harmonizes the large range of presently available manual protocols. In the near future, regionally differentiated automated methods will become available together with an appropriate statistical model, such as multivariate analysis of deformation fields, or techniques such as cortical-thickness measurements that yield a meaningful metrics for the detection of treatment effects. More advanced imaging protocols, including DTI, DSI, and functional MRI, are presently being used in monocenter and first multicenter studies. In the future these techniques will be relevant for the risk stratification in phase IIa type studies (small proof-of-concept trials). By contrast, the application of the broader established structural imaging biomarkers, such as hippocampus volume, for risk stratification and as surrogate end point is already today part of many clinical trial protocols. However, clinical care will also be affected by these new technologies. Radiologic expert centers already offer “dementia screening” for well-off middle-aged people who undergo an MRI scan with subsequent automated, typically VBM-based analysis, and determination of z-score deviation from a matched control cohort. Next-generation scanner software will likely include radiologic expert systems for automated segmentation, deformation-based morphometry, and multivariate analysis of anatomic MRI scans for the detection of a typical AD pattern. As these developments will start to change medical practice, first for selected subject groups that can afford this type of screening but later eventually also for other cohorts, clinicians must become aware of the potentials and limitations of these technologies. It is decidedly unclear to date how a middle-aged cognitively intact subject with a seemingly AD-positive MRI scan should be clinically advised. There is no evidence for individual risk prediction and even less for specific treatments. Thus, the development of preclinical diagnostic imaging poses not only technical but also ethical problems that must be critically discussed on the basis of profound knowledge. From a neurobiological point of view, the main determinants of cognitive impairment in AD are the density of synapses and neurons in distributed cortical and subcortical networks. MRI-based measures of regional gray matter volume and associated multivariate analysis techniques of regional interactions of gray matter densities provide insight into the onset and temporal dynamics of cortical atrophy as a close proxy for regional neuronal loss and a basis of functional impairment in specific neuronal networks. From the clinical point of view, clinicians must bear in mind that patients do not suffer from hippocampus atrophy or disconnection but from memory impairment, and that dementia screening in asymptomatic subjects should not be used outside of clinical studies.

Structural MRI in frontotemporal dementia: comparisons between hippocampal volumetry, tensor-based morphometry and voxel-based morphometry

BACKGROUND

MRI is an important clinical tool for diagnosing dementia-like diseases such as Frontemporal Dementia (FTD). However there is a need to develop more accurate and standardized MRI analysis methods.

OBJECTIVE

To compare FTD with Alzheimer's Disease (AD) and Mild Cognitive Impairment (MCI) with three automatic MRI analysis methods - Hippocampal Volumetry (HV), Tensor-based Morphometry (TBM) and Voxel-based Morphometry (VBM), in specific regions of interest in order to determine the highest classification accuracy.

METHODS

Thirty-seven patients with FTD, 46 patients with AD, 26 control subjects, 16 patients with progressive MCI (PMCI) and 48 patients with stable MCI (SMCI) were examined with HV, TBM for shape change, and VBM for gray matter density. We calculated the Correct Classification Rate (CCR), sensitivity (SS) and specificity (SP) between the study groups.

RESULTS

We found unequivocal results differentiating controls from FTD with HV (hippocampus left side) (CCR = 0.83; SS = 0.84; SP = 0.80), with TBM (hippocampus and amygdala (CCR = 0.80/SS = 0.71/SP = 0.94), and with VBM (all the regions studied, especially in lateral ventricle frontal horn, central part and occipital horn) (CCR = 0.87/SS = 0.81/SP = 0.96). VBM achieved the highest accuracy in differentiating AD and FTD (CCR = 0.72/SS = 0.67/SP = 0.76), particularly in lateral ventricle (frontal horn, central part and occipital horn) (CCR = 0.73), whereas TBM in superior frontal gyrus also achieved a high accuracy (CCR = 0.71/SS = 0.68/SP = 0.73). TBM resulted in low accuracy (CCR = 0.62) in the differentiation of AD from FTD using all regions of interest, with similar results for HV (CCR = 0.55).

CONCLUSION

Hippocampal atrophy is present not only in AD but also in FTD. Of the methods used, VBM achieved the highest accuracy in its ability to differentiate between FTD and AD.

Voxel and surface-based topography of memory and executive deficits in mild cognitive impairment and Alzheimer's disease

Mild cognitive impairment (MCI) and Alzheimer's disease (AD) are associated with a progressive loss of cognitive abilities. In the present report, we assessed the relationship of memory and executive function with brain structure in a sample of 810 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants, including 188 AD, 396 MCI, and 226 healthy older adults (HC). Composite scores of memory (ADNI-Mem) and executive function (ADNI-Exec) were generated by applying modern psychometric theory to item-level data from ADNI's neuropsychological battery. We performed voxel-based morphometry (VBM) and surface-based association (SurfStat) analyses to evaluate relationships of ADNI-Mem and ADNI-Exec with grey matter (GM) density and cortical thickness across the whole brain in the combined sample and within diagnostic groups. We observed strong associations between ADNI-Mem and medial and lateral temporal lobe atrophy. Lower ADNI-Exec scores were associated with advanced GM and cortical atrophy across broadly distributed regions, most impressively in the bilateral parietal and temporal lobes. We also evaluated ADNI-Exec adjusted for ADNI-Mem, and found associations with GM density and cortical thickness primarily in the bilateral parietal, temporal, and frontal lobes. Within-group analyses suggest these associations are strongest in patients with MCI and AD. The present study provides insight into the spatially unbiased associations between brain atrophy and memory and executive function, and underscores the importance of structural brain changes in early cognitive decline.

Amyloid positron emission tomography with (18)F-flutemetamol and structural magnetic resonance imaging in the classification of mild cognitive impairment and Alzheimer's disease

OBJECTIVE

To evaluate the contributions of amyloid-positive (Am+) and medial temporal atrophy-positive (MTA+) scans to the diagnostic classification of prodromal and probable Alzheimer's disease (AD).

METHODS

(18)F-flutemetamol-labeled amyloid positron emission tomography (PET) and magnetic resonance imaging (MRI) were used to classify 10 young normal, 15 elderly normal, 20 amnestic mild cognitive impairment (aMCI), and 27 AD subjects. MTA+ status was determined using a cut point derived from a previous study, and Am+ status was determined using a conservative and liberal cut point.

RESULTS

The rates of MRI scans with positive results among young normal, elderly normal, aMCI, and AD subjects were 0%, 20%, 75%, and 82%, respectively. Using conservative cut points, the rates of Am+ scans for these same groups of subjects were 0%, 7%, 50%, and 93%, respectively, with the aMCI group showing the largest discrepancy between Am+ and MTA+ scans. Among aMCI cases, 80% of Am+ subjects were also MTA+, and 70% of amyloid-negative (Am-) subjects were MTA+. The combination of amyloid PET and MTA data was additive, with an overall correct classification rate for aMCI of 86%, when a liberal cut point (standard uptake value ratio = 1.4) was used for amyloid positivity.

INTERPRETATION

(18)F-flutemetamol PET and structural MRI provided additive information in the diagnostic classification of aMCI subjects, suggesting an amyloid-independent neurodegenerative component among aMCI subjects in this sample.

A longitudinal study of atrophy in amnestic mild cognitive impairment and normal aging revealed by cortical thickness

In recent years, amnestic mild cognitive impairment (aMCI) has attracted significant attention as an indicator of high risk for Alzheimer's disease. An understanding of the pathology of aMCI may benefit the development of effective clinical treatments for dementia. In this work, we measured the cortical thickness of 109 aMCI subjects and 99 normal controls (NC) twice over two years. The longitudinal changes and the cross-sectional differences between the two types of participants were explored using the vertex thickness values. The thickness of the cortex in aMCI was found significantly reduced in both longitudinal and between-group comparisons, mainly in the temporal lobe, superolateral parietal lobe and some regions of the frontal cortices. Compared to NC, the aMCI showed a significantly high atrophy rate in the left lateral temporal lobe and left parahippocampal gyrus over two years. Additionally, a significant positive correlation between brain atrophy and the decline of Mini-Mental State Examination (MMSE) scores was also found in the left superior and left middle temporal gyrus in aMCI. These findings demonstrated specific longitudinal spatial patterns of cortical atrophy in aMCI and NC. The higher atrophy rate in aMCI might be responsible for the accelerated functional decline in the aMCI progression process.

Dementia and Alzheimer's disease: a new direction.The 2010 Jay L. Foster Memorial Lecture

BACKGROUND

The modern era of Alzheimer's disease (AD) research began in the early 1980s with the establishment of AD research centers and expanded research programs at the National Institute on Aging.

METHODS

Over the past 30 years, there has been success in defining criteria for AD and dementia, association of important genetic disorders related to premature dementia in families, the association of apolipoprotein-E(4), and measurement of incidence and prevalence and selected risk factors. However, prevention and treatment have been elusive.

RESULTS

The development of new technologies, especially magnetic resonance imaging, positron emission tomography to measure amyloid in vivo in the brain and glucose metabolism, cerebrospinal fluid examination, better genetic markers, large-scale longitudinal epidemiology studies, and preventive clinical trials has rapidly begun a new era of research that offers opportunities to better understand etiology, that is, determinants of amyloid biology in the brain, neurofibrillary tangles, synaptic loss, and dementia.

CONCLUSIONS

There are three major hypotheses related to dementia: amyloid deposition and secondary synaptic loss as a unique disease, vascular injury, and "aging." New research must be hypothesis-driven and lead to testable approaches for treatment and prevention.

Predicting future clinical changes of MCI patients using longitudinal and multimodal biomarkers

Accurate prediction of clinical changes of mild cognitive impairment (MCI) patients, including both qualitative change (i.e., conversion to Alzheimer's disease (AD)) and quantitative change (i.e., cognitive scores) at future time points, is important for early diagnosis of AD and for monitoring the disease progression. In this paper, we propose to predict future clinical changes of MCI patients by using both baseline and longitudinal multimodality data. To do this, we first develop a longitudinal feature selection method to jointly select brain regions across multiple time points for each modality. Specifically, for each time point, we train a sparse linear regression model by using the imaging data and the corresponding clinical scores, with an extra 'group regularization' to group the weights corresponding to the same brain region across multiple time points together and to allow for selection of brain regions based on the strength of multiple time points jointly. Then, to further reflect the longitudinal changes on the selected brain regions, we extract a set of longitudinal features from the original baseline and longitudinal data. Finally, we combine all features on the selected brain regions, from different modalities, for prediction by using our previously proposed multi-kernel SVM. We validate our method on 88 ADNI MCI subjects, with both MRI and FDG-PET data and the corresponding clinical scores (i.e., MMSE and ADAS-Cog) at 5 different time points. We first predict the clinical scores (MMSE and ADAS-Cog) at 24-month by using the multimodality data at previous time points, and then predict the conversion of MCI to AD by using the multimodality data at time points which are at least 6-month ahead of the conversion. The results on both sets of experiments show that our proposed method can achieve better performance in predicting future clinical changes of MCI patients than the conventional methods.

CSF VILIP-1 predicts rates of cognitive decline in early Alzheimer disease

OBJECTIVE

Measures of neuronal damage/dysfunction are likely good surrogates for disease progression in Alzheimer disease (AD). CSF markers of neuronal injury may offer utility in predicting disease progression and guiding prognostic and outcome assessments in therapeutic trials. Visinin-like protein-1 (VILIP-1) has demonstrated potential utility as a marker of neuronal injury. We here investigate the utility of VILIP-1 and VILIP-1/Aβ42 in predicting rates of cognitive decline in early AD.

METHODS

Individuals with a clinical diagnosis of very mild or mild AD (n = 60) and baseline CSF measures of VILIP-1, tau, p-tau181, and Aβ42 were followed longitudinally for an average of 2.6 years. Annual assessments included the Clinical Dementia Rating (CDR), CDR-sum of boxes (CDR-SB), and global composite scores. Mixed linear models assessed the ability of CSF biomarker measures to predict rates of cognitive decline over time.

RESULTS

Baseline CSF VILIP-1 and VILIP-1/Aβ42 levels predicted rates of future decline in CDR-SB and global composite scores over the follow-up period. Individuals with CSF VILIP-1 ≥560 pg/mL (corresponding to the upper tercile) progressed much more rapidly in CDR-SB (1.61 boxes/year; p = 0.0077) and global scores (-0.53 points/year; p = 0.0002) than individuals with lower values (0.85 boxes/year and -0.15 points/year, respectively) over the follow-up period. CSF tau, p-tau181, tau/Aβ42, and p-tau181/Aβ42 also predicted more rapid cognitive decline in CDR-SB and global scores over time.

CONCLUSION

These findings suggest that CSF VILIP-1 and VILIP-1/Aβ42 predict rates of global cognitive decline similarly to tau and tau/Aβ42, and may be useful CSF surrogates for neurodegeneration in early AD.

Combining structural brain changes improves the prediction of Alzheimer's disease and mild cognitive impairment

BACKGROUND

Several structural brain changes have been associated with Alzheimer's disease (AD). This study investigated the prediction of AD by combining multiple brain changes with the hallmark medial temporal lobe atrophy (MTA).

METHODS

High-resolution magnetic resonance imaging (MRI) data of people with mild AD (n = 39), mild cognitive impairment (MCI; n = 82), and of healthy controls (HC; n = 58) were obtained at baseline. Among these people, 26 AD, 53 MCI, and 46 HC subjects had 24-month follow-up MRI scans. Bilateral MTA was evaluated using a medial temporal lobe atrophy scale (MTAS). Common changes in the aging brain were summarized using a brain atrophy and lesion index (BALI). The performance of the MTAS, BALI, and a score combining both, using a logistic regression model, were assessed.

RESULTS

The MTAS and BALI scores were closely correlated (r(2) > 0.56); each differed between the diagnostic groups. Having an unfavorable MTAS score was associated with an increased risk of MCI-AD conversion (OR = 3.71, p = 0.039), adjusted for age, sex, and education; having an unfavorable BALI score marginally contributed to such risks (OR = 4.08, p = 0.080). Combining MTAS and BALI components resulted in a greater OR (8.99, p = 0.007) and an improved predictive accuracy (75.9%, p = 0.002).

CONCLUSION

Multiple structural changes have an additive effect on AD. The data support potential future roles of combining multiple coexisting structural changes to benefit AD diagnosis, progression monitoring, and/or treatment effect evaluation.

A comparison of screening tools for the assessment of mild cognitive impairment: preliminary findings

We report a pilot investigation into the utility of screening tools in Mild Cognitive Impairment (MCI). The Addenbrooke's Cognitive Examination-Revised (ACE-R), Montreal Cognitive Assessment (MoCA) and the novel Computer-Administered Neuropsychological Screen for Mild Cognitive Impairment (CANS-MCI) were administered to 20 elderly controls and 15 MCI cases. Non-parametric Mann-Whitney U-tests showed significant differences between groups (p < .0001) on the CANS-MCI and MoCA. The ACE-R and MoCA total scores showed high sensitivity (90%) to MCI. Area under the curve was consistently significant in discriminating controls and MCI for memory scores across all screening instruments. A useful profile of quantitative and qualitative information pertaining to cognitive functioning in MCI can be obtained with the MoCA, ACE-R, and CANS-MCI.

Constrained sparse functional connectivity networks for MCI classification

Mild cognitive impairment (MCI) is difficult to diagnose due to its subtlety. Recent emergence of advanced network analysis techniques utilizing resting-state functional Magnetic Resonance Imaging (rs-fMRI) has made the understanding of neurological disorders more comprehensively at a whole-brain connectivity level. However, inferring effective brain connectivity from fMRI data is a challenging task, particularly when the ultimate goal is to obtain good control-patient classification performance. Incorporating sparsity into connectivity modeling can potentially produce results that are biologically more meaningful since most biologically networks are formed by a relatively few number of connections. However, this constraint, when applied at an individual level, will degrade classification performance due to inter-subject variability. To address this problem, we consider a constrained sparse linear regression model associated with the least absolute shrinkage and selection operator (LASSO). Specifically, we introduced sparsity into brain connectivity via l1-norm penalization, and ensured consistent non-zero connections across subjects via l2-norm penalization. Our results demonstrate that the constrained sparse network gives better classification performance than the conventional correlation-based network, indicating its greater sensitivity to early stage brain pathologies.

Probable Alzheimer's disease patients presenting as "focal temporal lobe dysfunction" show a slow rate of cognitive decline

Several authors have recently shown that anterograde amnesia is often associated with semantic memory impairment in amnesic MCI patients. Similarly, after the MCI condition, some patients who convert to Alzheimer's disease (AD) show the classic onset (cAD) characterized by the impairment of memory and executive functions, whereas other AD patients show isolated defects of episodic and semantic memory without deficits in other cognitive domains. The latter have been considered an AD variant characterized by 'focal Temporal Lobe Dysfunction' (TLD). The aim of the present study was to assess the differences in disease progression between cAD and TLD. For this purpose a continuous series of newly diagnosed probable AD patients presenting as cAD (n = 30) and TLD (n = 25), matched for severity, and 65 healthy controls underwent a comprehensive neuropsychological evaluation at baseline; TLD and cAD were re-evaluated at a 24-month follow-up. At follow-up, TLD patients showed no significant worsening of cognitive functions, whereas cAD subjects displayed a significant worsening in all explored cognitive domains. In conclusion, our results confirm that probable AD presenting as TLD represents a specific onset of AD characterized by a slower rate of progression.

Generative-discriminative basis learning for medical imaging

This paper presents a novel dimensionality reduction method for classification in medical imaging. The goal is to transform very high-dimensional input (typically, millions of voxels) to a low-dimensional representation (small number of constructed features) that preserves discriminative signal and is clinically interpretable. We formulate the task as a constrained optimization problem that combines generative and discriminative objectives and show how to extend it to the semi-supervised learning (SSL) setting. We propose a novel large-scale algorithm to solve the resulting optimization problem. In the fully supervised case, we demonstrate accuracy rates that are better than or comparable to state-of-the-art algorithms on several datasets while producing a representation of the group difference that is consistent with prior clinical reports. Effectiveness of the proposed algorithm for SSL is evaluated with both benchmark and medical imaging datasets. In the benchmark datasets, the results are better than or comparable to the state-of-the-art methods for SSL. For evaluation of the SSL setting in medical datasets, we use images of subjects with mild cognitive impairment (MCI), which is believed to be a precursor to Alzheimer's disease (AD), as unlabeled data. AD subjects and normal control (NC) subjects are used as labeled data, and we try to predict conversion from MCI to AD on follow-up. The semi-supervised extension of this method not only improves the generalization accuracy for the labeled data (AD/NC) slightly but is also able to predict subjects which are likely to converge to AD.

Prediction of MCI to AD conversion, via MRI, CSF biomarkers, and pattern classification

Magnetic resonance imaging (MRI) patterns were examined together with cerebrospinal fluid (CSF) biomarkers in serial scans of Alzheimer's Disease Neuroimaging Initiative (ADNI) participants with mild cognitive impairment (MCI). The SPARE-AD score, summarizing brain atrophy patterns, was tested as a predictor of short-term conversion to Alzheimer's disease (AD). MCI individuals that converted to AD (MCI-C) had mostly positive baseline SPARE-AD (Spatial Pattern of Abnormalities for Recognition of Early AD) and atrophy in temporal lobe gray matter (GM) and white matter (WM), posterior cingulate/precuneous, and insula. MCI individuals that converted to AD had mostly AD-like baseline CSF biomarkers. MCI nonconverters (MCI-NC) had mixed baseline SPARE-AD and CSF values, suggesting that some MCI-NC subjects may later convert. Those MCI-NC with most negative baseline SPARE-AD scores (normal brain structure) had significantly higher baseline Mini Mental State Examination (MMSE) scores (28.67) than others, and relatively low annual rate of Mini Mental State Examination decrease (-0.25). MCI-NC with midlevel baseline SPARE-AD displayed faster annual rates of SPARE-AD increase (indicating progressing atrophy). SPARE-AD and CSF combination improved prediction over individual values. In summary, both SPARE-AD and CSF biomarkers showed high baseline sensitivity, however, many MCI-NC had abnormal baseline SPARE-AD and CSF biomarkers. Longer follow-up will elucidate the specificity of baseline measurements.

Alliance for aging research AD biomarkers work group: structural MRI

Biomarkers of Alzheimer's disease (AD) are increasingly important. All modern AD therapeutic trials employ AD biomarkers in some capacity. In addition, AD biomarkers are an essential component of recently updated diagnostic criteria for AD from the National Institute on Aging--Alzheimer's Association. Biomarkers serve as proxies for specific pathophysiological features of disease. The 5 most well established AD biomarkers include both brain imaging and cerebrospinal fluid (CSF) measures--cerebrospinal fluid Abeta and tau, amyloid positron emission tomography (PET), fluorodeoxyglucose (FDG) positron emission tomography, and structural magnetic resonance imaging (MRI). This article reviews evidence supporting the position that MRI is a biomarker of neurodegenerative atrophy. Topics covered include methods of extracting quantitative and semiquantitative information from structural MRI; imaging-autopsy correlation; and evidence supporting diagnostic and prognostic value of MRI measures. Finally, the place of MRI in a hypothetical model of temporal ordering of AD biomarkers is reviewed.

Hippocampal subregions are differentially affected in the progression to Alzheimer's disease

Atrophy within the hippocampus (HP) as measured by magnetic resonance imaging (MRI) is a promising biomarker for the progression to Alzheimer's disease (AD). Subregions of the HP along the longitudinal axis have been found to demonstrate unique function, as well as undergo differential changes in the progression to AD. Little is known of relationships between such HP subregions and other potential biomarkers, such as neuropsychological (NP), genetic, and cerebral spinal fluid (CSF) beta amyloid and tau measures. The purpose of this study was to subdivide the hippocampus to determine how the head, body, and tail were affected in normal control, mild cognitively impaired, and AD subjects, and investigate relationships with HP subregions and other potential biomarkers. MRI scans of 120 participants of the Alzheimer's Disease Neuroimaging Initiative were processed using FreeSurfer, and the HP was subdivided using 3D Slicer. Each subregion was compared among groups, and correlations were used to determine relationships with NP, genetic, and CSF measures. Results suggest that HP subregions are undergoing differential atrophy in AD, and demonstrate unique relationships with NP and CSF data. Discriminant function analyses revealed that these regions, when combined with NP and CSF measures, were able to classify by diagnostic group, and classify MCI subjects who would and would not progress to AD within 12 months.

Multi-modal multi-task learning for joint prediction of multiple regression and classification variables in Alzheimer's disease

Many machine learning and pattern classification methods have been applied to the diagnosis of Alzheimer's disease (AD) and its prodromal stage, i.e., mild cognitive impairment (MCI). Recently, rather than predicting categorical variables as in classification, several pattern regression methods have also been used to estimate continuous clinical variables from brain images. However, most existing regression methods focus on estimating multiple clinical variables separately and thus cannot utilize the intrinsic useful correlation information among different clinical variables. On the other hand, in those regression methods, only a single modality of data (usually only the structural MRI) is often used, without considering the complementary information that can be provided by different modalities. In this paper, we propose a general methodology, namely multi-modal multi-task (M3T) learning, to jointly predict multiple variables from multi-modal data. Here, the variables include not only the clinical variables used for regression but also the categorical variable used for classification, with different tasks corresponding to prediction of different variables. Specifically, our method contains two key components, i.e., (1) a multi-task feature selection which selects the common subset of relevant features for multiple variables from each modality, and (2) a multi-modal support vector machine which fuses the above-selected features from all modalities to predict multiple (regression and classification) variables. To validate our method, we perform two sets of experiments on ADNI baseline MRI, FDG-PET, and cerebrospinal fluid (CSF) data from 45 AD patients, 91 MCI patients, and 50 healthy controls (HC). In the first set of experiments, we estimate two clinical variables such as Mini Mental State Examination (MMSE) and Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-Cog), as well as one categorical variable (with value of 'AD', 'MCI' or 'HC'), from the baseline MRI, FDG-PET, and CSF data. In the second set of experiments, we predict the 2-year changes of MMSE and ADAS-Cog scores and also the conversion of MCI to AD from the baseline MRI, FDG-PET, and CSF data. The results on both sets of experiments demonstrate that our proposed M3T learning scheme can achieve better performance on both regression and classification tasks than the conventional learning methods.

Neuroanatomic changes and their association with cognitive decline in mild cognitive impairment: a meta-analysis

Mild cognitive impairment (MCI) is an acquired syndrome characterised by cognitive decline not affecting activities of daily living. Using a quantitative meta-analytic approach, we aimed to identify consistent neuroanatomic correlates of MCI and how they are related to cognitive dysfunction. The meta-analysis enrols 22 studies, involving 917 MCI (848 amnestic MCI) patients and 809 healthy controls. Only studies investigating local changes in grey matter and reporting whole-brain results in stereotactic coordinates were included and analysed using the activation likelihood estimation approach. Probabilistic cytoarchitectonic maps were used to compare the localization of the obtained significant effects to histological areas. A correlation between the probability of grey matter changes and cognitive performance of MCI patients was performed. In MCI patients, the meta-analysis revealed three significant clusters of convergent grey matter atrophy, which were mainly situated in the bilateral amygdala and hippocampus, extending to the left medial temporal pole and thalamus, as well as in the bilateral precuneus. A sub-analysis in only amnestic MCI revealed a similar pattern. A voxel-wise analysis revealed a correlation between grey matter reduction and cognitive decline in the right hippocampus and amygdala as well as in the left thalamus. This study provides convergent evidence of a distinct neuroanatomical pattern in MCI. The correlation analysis with cognitive-mnestic decline further highlights the impact of limbic structures and the linkage with data from a functional neuroimaging database provides additional insight into underlying functions. Although different pathologies are underlying MCI, the observed neuroanatomical pattern of structural changes may reflect the common clinical denominator of cognitive impairment.

Pattern Analysis in Neuroimaging: Beyond Two-Class Categorization

One of the many advantages of multivariate pattern recognition approaches over conventional mass-univariate group analysis using voxel-wise statistical tests is their potential to provide highly sensitive and specific markers of diseases on an individual basis. However, a vast majority of imaging problems addressed by pattern recognition are viewed from the perspective of a two-class classification. In this article, we provide a summary of selected works that propose solutions to biomedical problems where the widely-accepted classification paradigm is not appropriate. These pattern recognition approaches address common challenges in many imaging studies: high heterogeneity of populations and continuous progression of diseases. We focus on diseases associated with aging and propose that clustering-based approaches may be more suitable for disentanglement of the underlying heterogeneity, while high-dimensional pattern regression methodology is appropriate for prediction of continuous and gradual clinical progression from magnetic resonance brain images.

Identifying cognitively healthy elderly individuals with subsequent memory decline by using automated MR temporoparietal volumes

PURPOSE

To determine whether automated temporoparietal brain volumes can be used to accurately predict future memory decline among a multicenter cohort of cognitively healthy elderly individuals.

MATERIALS AND METHODS

The study was approved by the institutional review board at each site and was HIPAA compliant, with written consent obtained from all participants. One hundred forty-nine cognitively healthy study participants were recruited through the Alzheimer's Disease Neuroimaging Initiative and underwent a standardized baseline 1.5-T magnetic resonance (MR) imaging examination, as well as neuropsychological assessment at baseline and after 2 years of follow-up. A composite memory score for the 2-year change in the results of two delayed-recall tests was calculated, and memory decline was defined as a composite score that was at least 1 standard deviation below the group mean score. The predictive accuracy of the brain volumes was estimated by using areas under receiver operating characteristic curves and was further assessed by using leave-one-out cross validation.

RESULTS

Use of the most accurate region model, which included the hippocampus; parahippocampal gyrus; amygdala; superior, middle, and inferior temporal gyri; superior parietal lobe; and posterior cingulate gyrus, resulted in a fitted accuracy of 94% and a cross-validated accuracy of 81%.

CONCLUSION

Study results indicate that automated temporal and parietal volumes can be used to identify with high accuracy cognitively healthy individuals who are at risk for future memory decline. Further validation of this predictive model in a new cohort is required.

Thickness in entorhinal and subicular cortex predicts episodic memory decline in mild cognitive impairment

Identifying subjects with mild cognitive impairment (MCI) most likely to decline in cognition over time is a major focus in Alzheimer's disease (AD) research. Neuroimaging biomarkers that predict decline would have great potential for increasing the efficacy of early intervention. In this study, we used high-resolution MRI, combined with a cortical unfolding technique to increase visibility of the convoluted medial temporal lobe (MTL), to assess whether gray matter thickness in subjects with MCI correlated to decline in cognition over two years. We found that thickness in the entorhinal (ERC) and subicular (Sub) cortices of MCI subjects at initial assessment correlated to change in memory encoding over two years (ERC: r = 0.34; P = .003) and Sub (r = 0.26; P = .011) but not delayed recall performance. Our findings suggest that aspects of memory performance may be differentially affected in the early stages of AD. Given the MTL's involvement in early stages of neurodegeneration in AD, clarifying the relationship of these brain regions and the link to resultant cognitive decline is critical in understanding disease progression.

The search for neuroimaging biomarkers of Alzheimer's disease with advanced MRI techniques

The aim of this review is to examine the recent literature on using advanced magnetic resonance imaging (MRI) techniques for finding neuroimaging biomarkers that are sensitive to the detection of risks for Alzheimer's disease (AD). Since structural MRI techniques, such as brain structural volumetry and voxel-based morphometry (VBM), have been widely used for AD studies and extensively reviewed, we will only briefly touch on the topics of volumetry and morphometry. The focus of the current review is about the more recent developments in the search for AD neuroimaging biomarkers with functional MRI (fMRI), resting-state functional connectivity MRI (fcMRI), diffusion tensor imaging (DTI), arterial spin-labeling (ASL), and magnetic resonance spectroscopy (MRS).

Structural brain changes in normal individuals with a maternal history of Alzheimer's

Having a parent affected with late-onset Alzheimer's disease (LOAD) is a major risk factor for developing the disease among cognitively normal (NL) individuals. This magnetic resonance imaging (MRI) study examines whether NL with a LOAD-affected parent show preclinical brain atrophy, and whether there are parent-of-origin effects. Voxel-based morphometry (VBM) on Statistical parametric mapping (SPM8) was used to examine volumetric T1-MRI scans of 60 late-middle-aged NL subjects, divided into 3 size-matched, demographically balanced groups of 20 subjects each, including NL with a maternal (FHm), paternal (FHp), or negative family history (FH-) of LOAD. There were no group differences for clinical and neuropsychological measures, and ApoE status. On VBM, FHm showed reduced gray matter volumes (GMV) in frontal, parietal, and temporal cortices and precuneus as compared with FH-, and in precuneus compared with FHp (p < 0.05, family-wise error [FWE]-corrected). Results remained significant controlling for age, gender, education, ApoE, and total intracranial volume. No differences were observed between FHp and FH- in any regions. NL FHm showed reduced GMV in LOAD-affected brain regions compared with FH- and FHp, indicating higher risk for Alzheimer's disease. Our findings support the use of regional brain atrophy as a preclinical biomarker for LOAD among at-risk individuals.

An examination of an enhancing effect of music on attentional abilities in older persons with mild cognitive impairment

While the effect of listening to music on cognitive abilities is highly debated, studies reporting an enhancing effect of music in elderly populations appear to be more consistent. In this study, the effects of listening to music on attention in groups of cognitively normal older adults and those with mild cognitive impairment were considered. Participants were exposed to both a music and silence condition, and after each condition performed Digit Span and Coding tasks which require attention for maximal performance. The hypothesis that listening to music, compared to a silence condition, enhances performance was not supported for either group. Various explanations for these findings are considered.

Multimodal classification of Alzheimer's disease and mild cognitive impairment

Effective and accurate diagnosis of Alzheimer's disease (AD), as well as its prodromal stage (i.e., mild cognitive impairment (MCI)), has attracted more and more attention recently. So far, multiple biomarkers have been shown to be sensitive to the diagnosis of AD and MCI, i.e., structural MR imaging (MRI) for brain atrophy measurement, functional imaging (e.g., FDG-PET) for hypometabolism quantification, and cerebrospinal fluid (CSF) for quantification of specific proteins. However, most existing research focuses on only a single modality of biomarkers for diagnosis of AD and MCI, although recent studies have shown that different biomarkers may provide complementary information for the diagnosis of AD and MCI. In this paper, we propose to combine three modalities of biomarkers, i.e., MRI, FDG-PET, and CSF biomarkers, to discriminate between AD (or MCI) and healthy controls, using a kernel combination method. Specifically, ADNI baseline MRI, FDG-PET, and CSF data from 51AD patients, 99 MCI patients (including 43 MCI converters who had converted to AD within 18 months and 56 MCI non-converters who had not converted to AD within 18 months), and 52 healthy controls are used for development and validation of our proposed multimodal classification method. In particular, for each MR or FDG-PET image, 93 volumetric features are extracted from the 93 regions of interest (ROIs), automatically labeled by an atlas warping algorithm. For CSF biomarkers, their original values are directly used as features. Then, a linear support vector machine (SVM) is adopted to evaluate the classification accuracy, using a 10-fold cross-validation. As a result, for classifying AD from healthy controls, we achieve a classification accuracy of 93.2% (with a sensitivity of 93% and a specificity of 93.3%) when combining all three modalities of biomarkers, and only 86.5% when using even the best individual modality of biomarkers. Similarly, for classifying MCI from healthy controls, we achieve a classification accuracy of 76.4% (with a sensitivity of 81.8% and a specificity of 66%) for our combined method, and only 72% even using the best individual modality of biomarkers. Further analysis on MCI sensitivity of our combined method indicates that 91.5% of MCI converters and 73.4% of MCI non-converters are correctly classified. Moreover, we also evaluate the classification performance when employing a feature selection method to select the most discriminative MR and FDG-PET features. Again, our combined method shows considerably better performance, compared to the case of using an individual modality of biomarkers.

Increased fMRI signal with age in familial Alzheimer's disease mutation carriers

Although many Alzheimer's disease (AD) patients have a family history of the disease, it is rarely inherited in a predictable way. Functional magnetic resonance imaging (fMRI) studies of nondemented adults carrying familial AD mutations provide an opportunity to prospectively identify brain differences associated with early AD-related changes. We compared fMRI activity of 18 nondemented autosomal dominant AD mutation carriers with fMRI activity in eight of their noncarrier relatives as they performed a novelty encoding task in which they viewed novel and repeated images. Because age of disease onset is relatively consistent within families, we also correlated fMRI activity with subjects' distance from the median age of diagnosis for their family. Mutation carriers did not show significantly different voxelwise fMRI activity from noncarriers as a group. However, as they approached their family age of disease diagnosis, only mutation carriers showed increased fMRI activity in the fusiform and middle temporal gyri. This suggests that during novelty encoding, increased fMRI activity in the temporal lobe may relate to incipient AD processes.

Automatic classification of patients with Alzheimer's disease from structural MRI: a comparison of ten methods using the ADNI database

Recently, several high dimensional classification methods have been proposed to automatically discriminate between patients with Alzheimer's disease (AD) or mild cognitive impairment (MCI) and elderly controls (CN) based on T1-weighted MRI. However, these methods were assessed on different populations, making it difficult to compare their performance. In this paper, we evaluated the performance of ten approaches (five voxel-based methods, three methods based on cortical thickness and two methods based on the hippocampus) using 509 subjects from the ADNI database. Three classification experiments were performed: CN vs AD, CN vs MCIc (MCI who had converted to AD within 18 months, MCI converters - MCIc) and MCIc vs MCInc (MCI who had not converted to AD within 18 months, MCI non-converters - MCInc). Data from 81 CN, 67 MCInc, 39 MCIc and 69 AD were used for training and hyperparameters optimization. The remaining independent samples of 81 CN, 67 MCInc, 37 MCIc and 68 AD were used to obtain an unbiased estimate of the performance of the methods. For AD vs CN, whole-brain methods (voxel-based or cortical thickness-based) achieved high accuracies (up to 81% sensitivity and 95% specificity). For the detection of prodromal AD (CN vs MCIc), the sensitivity was substantially lower. For the prediction of conversion, no classifier obtained significantly better results than chance. We also compared the results obtained using the DARTEL registration to that using SPM5 unified segmentation. DARTEL significantly improved six out of 20 classification experiments and led to lower results in only two cases. Overall, the use of feature selection did not improve the performance but substantially increased the computation times.

Automated MR morphometry to predict Alzheimer's disease in mild cognitive impairment

PURPOSE

Prediction of progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD) is challenging but essential for early treatment. This study aims to investigate the use of hippocampal atrophy markers for the automatic detection of MCI converters and to compare the predictive value to manually obtained hippocampal volume and temporal horn width.

METHODS

A study was performed with 15 patients with Alzheimer and 18 patients with MCI (ten converted, eight remained stable in a 3-year follow-up) as well as 15 healthy subjects. MRI scans were obtained at baseline and evaluated with an automated system for scoring of hippocampal atrophy. The predictive value of the automated system was compared with manual measurements of hippocampal volume and temporal horn width in the same subjects.

RESULTS

The conversion to AD was correctly predicted in 77.8% of the cases (sensitivity 70%, specificity 87.5%) in the MCI group using automated morphometry and a plain linear classifier that was trained on the AD and healthy groups. Classification was improved by limiting analysis to the left cerebral hemisphere (accuracy 83.3%, sensitivity 70%, specificity 100%). The manual linear and volumetric approaches reached rates of 66.7% (40/100%) and 72.2% (60/87.5%), respectively.

CONCLUSION

The automatic approach fulfills many important preconditions for clinical application. Contrary to the manual approaches, it is not observer-dependent and reduces human resource requirements. Automated assessment may be useful for individual patient assessment and for predicting progression to dementia.

Biomarkers in translational research of Alzheimer's disease

The identification and characterization of amyloid-beta (Abeta) and tau as the main pathological substrates of Alzheimer's disease (AD) have driven many efforts in search for suitable biomarkers for AD. In the last decade, research in this area has focused on developing a better understanding of the principles that govern protein deposition, mechanisms that link aggregation to toxicity and neuronal death, and a better understanding of protein dynamics in brain tissue, interstitial fluid and CSF. While Abeta and tau represent the two key pathological mediators of disease, other aspects of this multifaceted disease (e.g. oxidative stress, calcium-mediated toxicity, and neuroinflammation) are being unraveled, with the hope to develop a more comprehensive approach in exploring disease mechanisms. This has not only expanded possible areas for disease-modifying therapies, but has also allowed the introduction of novel, and potentially useful, fluid and radiological markers for the presence and progression of AD pathology. There is no doubt that the identification of several fluid and imaging biomarkers that can reliably detect the early stages of AD will have great implications in the design of clinical trials, in the selection of homogenous research populations, and in the assessment of disease outcomes. Markers with good diagnostic specificity will aid researchers in differentiating individuals with preclinical and probable AD from individuals who do not have AD pathology or have other dementing disorders. Markers that change with disease progression may offer utility in assessing the rates of disease progression and the efficacy of potential therapeutic agents on AD pathology. For both of these purposes, CSF Abeta42, amyloid imaging, and CSF tau appear to be very good markers of the presence of AD pathology as well as predictive of who will progress from MCI to AD. Volumetric MRI is also good at separating individuals with MCI and AD from controls and is predictive of who will progress from MCI to AD. Perhaps the most important role biomarkers will have, and the most needed at this time, lies in the identification of individuals who are cognitively normal, and yet have evidence of AD pathology (i.e. preclinical AD). Such individuals, it appears, can be identified with CSF Abeta42, amyloid imaging, and CSF tau. Such individuals are the most likely to benefit from future disease modifying/prevention therapies as they become available, and therefore represent the population in which the field can make the biggest therapeutic impact.

Evidence of neurodegeneration in brains of older adults who do not yet fulfill MCI criteria

We sought to determine whether there are structural and metabolic changes in the brains of older adults with cognitive complaints yet who do not meet MCI criteria (i.e., preMCI). We compared the volumes of regional lobar gray matter (GM) and medial temporal lobe structures, including the hippocampus, entorhinal cortex (ERC), fusiform and parahippocampal gyri, and metabolite ratios from the posterior cingulate in individuals who had a Clinical Demetia Rating (CDR) of 0.5, but who did not meet MCI criteria (preMCI, N=17), patients with mild cognitive impairment (MCI, N=13), and cognitively normal controls (N=18). Controls had more ERC, fusiform, and frontal gray matter volume than preMCI and MCI subjects and greater parahippocampal volume and more posterior cingulate N-acetylaspartate (NAA)/myoinosotil (mI) than MCI. There were no significant differences between MCI and preMCI subjects on any of these measures. These findings suggest there are neurodegenerative changes in the brains of older adults who have cognitive complaints severe enough to qualify for CDR=0.5 yet show no deficits on formal neuropsychological testing. The results further support the hypothesis that detection of individuals with very mild forms of Alzheimer's disease (AD) may be facilitated by use of the CDR, which emphasizes changes in cognition over time within individuals rather than comparison with group norms.

Mapping brain morphological and functional conversion patterns in amnestic MCI: a voxel-based MRI and FDG-PET study

PURPOSE

To reveal the morphological and functional substrates of memory impairment and conversion to Alzheimer disease (AD) from the stage of amnestic mild cognitive impairment (aMCI).

METHODS

Brain MRI and FDG-PET were performed in 20 patients with aMCI and 12 controls at baseline. During a mean follow-up of about 2 years, 9 patients developed AD (converters), and 11 did not (nonconverters). All images were processed with SPM2. FDG-PET and segmented grey matter (GM) images were compared in: (1) converters versus controls, (2) nonconverters versus controls, and (3) converters versus nonconverters.

RESULTS

As compared to controls, converters showed lower GM density in the left parahippocampal gyrus and both thalami, and hypometabolism in the precuneus, posterior cingulate and superior parietal lobule in the left hemisphere. Hypometabolism was found in nonconverters as compared to controls in the left precuneus and posterior cingulated gyrus. As compared to nonconverters, converters showed significant hypometabolism in the left middle and superior temporal gyri.

CONCLUSION

The discordant topography between atrophy and hypometabolism reported in AD is already present at the aMCI stage. Posterior cingulate-precuneus hypometabolism seemed to be an early sign of memory deficit, whereas hypometabolism in the left temporal cortex marked the conversion to AD.