Search

IntroductionThe Centiloid scale was developed to harmonise the quantification of β-amyloid (Aβ) PET images across tracers, scanners, and processing pipelines. However, several groups have reported differences across tracers and scanners even after centiloid conversion. In this study, we aim to evaluate the impact of different pre and post-processing harmonisation steps on the robustness of longitudinal Centiloid data across three large international cohort studies.MethodsAll Aβ PET data in AIBL (N = 3315), ADNI (N = 3442) and OASIS3 (N = 1398) were quantified using the MRI-based Centiloid standard SPM pipeline and the PET-only pipeline CapAIBL. SUVR were converted into Centiloids using each tracer's respective transform. Global Aβ burden from pre-defined target cortical regions in Centiloid units were quantified for both raw PET scans and PET scans smoothed to a uniform 8 mm full width half maximum (FWHM) effective smoothness. For Florbetapir, we assessed the performance of using both the standard Whole Cerebellum (WCb) and a composite white matter (WM)+WCb reference region. Additionally, our recently proposed quantification based on Non-negative Matrix Factorisation (NMF) was applied to all spatially and SUVR normalised images. Correlation with clinical severity measured by the Mini-Mental State Examination (MMSE) and effect size, as well as tracer agreement in 11C-PiB-18F-Florbetapir pairs and longitudinal consistency were evaluated.ResultsThe smoothing to a uniform resolution partially reduced longitudinal variability, but did not improve inter-tracer agreement, effect size or correlation with MMSE. Using a Composite reference region for 18F-Florbetapir improved inter-tracer agreement, effect size, correlation with MMSE, and longitudinal consistency. The best results were however obtained when using the NMF method which outperformed all other quantification approaches in all metrics used.ConclusionsFWHM smoothing has limited impact on longitudinal consistency or outliers. A Composite reference region including subcortical WM should be used for computing both cross-sectional and longitudinal Florbetapir Centiloid. NMF improves Centiloid quantification on all metrics examined.

BackgroundAlthough altered lipid metabolism has been extensively implicated in the pathogenesis of Alzheimer disease (AD) through cell biological, epidemiological, and genetic studies, the molecular mechanisms linking cholesterol and AD pathology are still not well understood and contradictory results have been reported. We have used a Mendelian randomization approach to dissect the causal nature of the association between circulating lipid levels and late onset AD (LOAD) and test the hypothesis that genetically raised lipid levels increase the risk of LOAD.Methods and findingsWe included 3,914 patients with LOAD, 1,675 older individuals without LOAD, and 4,989 individuals from the general population from six genome wide studies drawn from a white population (total n=10,578). We constructed weighted genotype risk scores (GRSs) for four blood lipid phenotypes (high-density lipoprotein cholesterol [HDL-c], low-density lipoprotein cholesterol [LDL-c], triglycerides, and total cholesterol) using well-established SNPs in 157 loci for blood lipids reported by Willer and colleagues (2013). Both full GRSs using all SNPs associated with each trait at pConclusionsGenetic predisposition to increased blood cholesterol and triglyceride lipid levels is not associated with elevated LOAD risk. The observed epidemiological associations between abnormal lipid levels and LOAD risk could therefore be attributed to the result of biological pleiotropy or could be secondary to LOAD. Limitations of this study include the small proportion of lipid variance explained by the GRS, biases in case-control ascertainment, and the limitations implicit to Mendelian randomization studies. Future studies should focus on larger LOAD datasets with longitudinal sampled peripheral lipid measures and other markers of lipid metabolism, which have been shown to be altered in LOAD. Please see later in the article for the Editors' Summary.

We present a method to discover discriminative brain metabolism patterns in [18F] fluorodeoxyglucose positron emission tomography (PET) scans, facilitating the clinical diagnosis of Alzheimer's disease. In the work, the term "pattern" stands for a certain brain region that characterizes a target group of patients and can be used for a classification as well as interpretation purposes. Thus, it can be understood as a so-called "region of interest (ROI)". In the literature, an ROI is often found by a given brain atlas that defines a number of brain regions, which corresponds to an anatomical approach. The present work introduces a semi-data-driven approach that is based on learning the characteristics of the given data, given some prior anatomical knowledge. A Gaussian Mixture Model (GMM) and model selection are combined to return a clustering of voxels that may serve for the definition of ROIs. Experiments on both an in-house dataset and data of the Alzheimer's Disease Neuroimaging Initiative (ADNI) suggest that the proposed approach arrives at a better diagnosis than a merely anatomical approach or conventional statistical hypothesis testing.

This paper describes a novel method to automatically segment the human brainstem into midbrain and pons, called labs: Landmark-based Automated Brainstem Segmentation. LABS processes high-resolution structural magnetic resonance images (MRIs) according to a revised landmark-based approach integrated with a thresholding method, without manual interaction.This method was first tested on morphological T1-weighted MRIs of 30 healthy subjects. Its reliability was further confirmed by including neurological patients (with Alzheimer's Disease) from the ADNI repository, in whom the presence of volumetric loss within the brainstem had been previously described. Segmentation accuracies were evaluated against expert-drawn manual delineation. To evaluate the quality of LABS segmentation we used volumetric, spatial overlap and distance-based metrics.The comparison between the quantitative measurements provided by LABS against manual segmentations revealed excellent results in healthy controls when considering either the midbrain (DICE measures higher that 0.9; Volume ratio around 1 and Hausdorff distance around 3) or the pons (DICE measures around 0.93; Volume ratio ranging 1.024-1.05 and Hausdorff distance around 2). Similar performances were detected for AD patients considering segmentation of the pons (DICE measures higher that 0.93; Volume ratio ranging from 0.97-0.98 and Hausdorff distance ranging 1.07-1.33), while LABS performed lower for the midbrain (DICE measures ranging 0.86-0.88; Volume ratio around 0.95 and Hausdorff distance ranging 1.71-2.15).Our study represents the first attempt to validate a new fully automated method for in vivo segmentation of two anatomically complex brainstem subregions. We retain that our method might represent a useful tool for future applications in clinical practice.

This paper presents a method for selecting Regions of Interest (ROI) in brain Magnetic Resonance Imaging (MRI) for diagnostic purposes, using statistical learning and vector quantization techniques. The proposed method models the distribution of GM and WM tissues grouping the voxels belonging to each tissue in ROIs associated to a specific neurological disorder. Tissue distribution of normal and abnormal images is modelled by a Self-Organizing map (SOM), generating a set of representative prototypes, and the receptive field (RF) of each SOM prototype defines a ROI. Moreover, the proposed method computes the relative importance of each ROI by means of its discriminative power. The devised method has been assessed using 818 images from the Alzheimer's disease Neuroimaging Initiative (ADNI) which were previously segmented through Statistical Parametric Mapping (SPM). The proposed algorithm was used over these images to parcel ROIs associated to the Alzheimer's Disease (AD). Additionally, this method can be used to extract a reduced set of discriminative features for classification, since it compresses discriminative information contained in the brain. Voxels marked by ROIs which were computed using the proposed method, yield classification results up to 90% of accuracy for controls (CN) and Alzheimer's disease (AD) patients, and 84% of accuracy for Mild Cognitive Impairment (MCI) and AD patients.

Genome-wide association studies (GWAS) have successfully identified several risk loci for Alzheimer's disease (AD). Nonetheless, these loci do not explain the entire susceptibility of the disease, suggesting that other genetic contributions remain to be identified. Here, we performed a meta-analysis combining data of 4,569 individuals (2,540 cases and 2,029 healthy controls) derived from three publicly available GWAS in AD and replicated a broad genomic region (>248,000 bp) associated with the disease near the APOE/TOMM40 locus in chromosome 19. To detect minor effect size contributions that could help to explain the remaining genetic risk, we conducted network-based pathway analyses either by extracting gene-wise p-values (GW), defined as the single strongest association signal within a gene, or calculated a more stringent gene-based association p-value using the extended Simes (GATES) procedure. Comparison of these strategies revealed that ontological sub-networks (SNs) involved in glutamate signaling were significantly overrepresented in AD (p

Accurate and robust brain extraction is a critical step in most neuroimaging analysis pipelines. In particular, for the large-scale multi-site neuroimaging studies involving a significant number of subjects with diverse age and diagnostic groups, accurate and robust extraction of the brain automatically and consistently is highly desirable. In this paper, we introduce population-specific probability maps to guide the brain extraction of diverse subject groups, including both healthy and diseased adult human populations, both developing and aging human populations, as well as non-human primates. Specifically, the proposed method combines an atlas-based approach, for coarse skull-stripping, with a deformable-surface-based approach that is guided by local intensity information and population-specific prior information learned from a set of real brain images for more localized refinement. Comprehensive quantitative evaluations were performed on the diverse large-scale populations of ADNI dataset with over 800 subjects (55 ∼ 90 years of age, multi-site, various diagnosis groups), OASIS dataset with over 400 subjects (18 ∼ 96 years of age, wide age range, various diagnosis groups), and NIH pediatrics dataset with 150 subjects (5 ∼ 18 years of age, multi-site, wide age range as a complementary age group to the adult dataset). The results demonstrate that our method consistently yields the best overall results across almost the entire human life span, with only a single set of parameters. To demonstrate its capability to work on non-human primates, the proposed method is further evaluated using a rhesus macaque dataset with 20 subjects. Quantitative comparisons with popularly used state-of-the-art methods, including BET, Two-pass BET, BET-B, BSE, HWA, ROBEX and AFNI, demonstrate that the proposed method performs favorably with superior performance on all testing datasets, indicating its robustness and effectiveness.

Alzheimer's disease (AD) has a strong propensity to run in families. However, the known risk genes excluding APOE are not clinically useful. In various complex diseases, gene studies have targeted rare alleles for unsolved heritability. Our study aims to elucidate previously unknown risk genes for AD by targeting rare alleles. We used data from five publicly available genetic studies from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and the database of Genotypes and Phenotypes (dbGaP). A total of 4,171 cases and 9,358 controls were included. The genotype information of rare alleles was imputed using 1,000 genomes. We performed gene-based analysis of rare alleles (minor allele frequency≤3%). The genome-wide significance level was defined as meta P

We investigated the neural basis of hallucinations Alzheimer's disease (AD) by applying voxel-based morphometry (VBM) to anatomical and functional data from the AD Neuroimaging Initiative.AD patients with hallucinations, based on the Neuropsychiatric Inventory (NPI-Q) (AD-hallu group; n = 39), were compared to AD patients without hallucinations matched for age, sex, educational level, handedness and MMSE (AD-c group; n = 39). Focal brain volume on MRI was analyzed and compared between the two groups according to the VBM method. We also performed voxel-level correlations between brain volume and hallucinations intensity. A similar paradigm was used for the PET analysis. "Core regions" (i.e. regions identified in both MRI and PET analyses, simply done by retaining the clusters obtained from the two analyses that are overlapping) were then determined.Regions with relative atrophy in association with hallucinations were: anterior part of the right insula, left superior frontal gyrus and lingual gyri. Regions with relative hypometabolism in association with hallucinations were a large right ventral and dorsolateral prefrontal area. "Core region" in association with hallucinations was the right anterior part of the insula. Correlations between intensity of hallucinations and brain volume were found in the right anterior insula, precentral gyrus, superior temporal gyrus, and left precuneus. Correlations between intensity of hallucinations and brain hypometabolism were found in the left midcingulate gyrus. We checked the neuropathological status and we found that the 4 patients autopsied in the AD-hallu group had the mixed pathology AD and Dementia with Lewy bodies (DLB).Neural basis of hallucinations in cognitive neurodegenerative diseases (AD or AD and DLB) include a right predominant anterior-posterior network, and the anterior insula as the core region. This study is coherent with the top-down/bottom-up hypotheses on hallucinations but also hypotheses of the key involvement of the anterior insula in hallucinations in cognitive neurodegenerative diseases.

To investigate whether APOE ε4 carriers have higher hippocampal atrophy rates than non-carriers in Alzheimer's disease (AD), mild cognitive impairment (MCI) and controls, and if so, whether higher hippocampal atrophy rates are still observed after adjusting for concurrent whole-brain atrophy rates.MRI scans from all available visits in ADNI (148 AD, 307 MCI, 167 controls) were used. MCI subjects were divided into "progressors" (MCI-P) if diagnosed with AD within 36 months or "stable" (MCI-S) if a diagnosis of MCI was maintained. A joint multi-level mixed-effect linear regression model was used to analyse the effect of ε4 carrier-status on hippocampal and whole-brain atrophy rates, adjusting for age, gender, MMSE and brain-to-intracranial volume ratio. The difference in hippocampal rates between ε4 carriers and non-carriers after adjustment for concurrent whole-brain atrophy rate was then calculated.Mean adjusted hippocampal atrophy rates in ε4 carriers were significantly higher in AD, MCI-P and MCI-S (p≤0.011, all tests) compared with ε4 non-carriers. After adjustment for whole-brain atrophy rate, the difference in mean adjusted hippocampal atrophy rate between ε4 carriers and non-carriers was reduced but remained statistically significant in AD and MCI-P.These results suggest that the APOE ε4 allele drives atrophy to the medial-temporal lobe region in AD.

Machine learning techniques, along with imaging markers extracted from structural magnetic resonance images, have been shown to increase the accuracy to differentiate patients with Alzheimer's disease (AD) from normal elderly controls. Several forms of anatomical features, such as cortical volume, shape, and thickness, have demonstrated discriminative capability. These approaches rely on accurate non-linear image transformation, which could invite several nuisance factors, such as dependency on transformation parameters and the degree of anatomical abnormality, and an unpredictable influence of residual registration errors. In this study, we tested a simple method to extract disease-related anatomical features, which is suitable for initial stratification of the heterogeneous patient populations often encountered in clinical data. The method employed gray-level invariant features, which were extracted from linearly transformed images, to characterize AD-specific anatomical features. The intensity information from a disease-specific spatial masking, which was linearly registered to each patient, was used to capture the anatomical features. We implemented a two-step feature selection for anatomic recognition. First, a statistic-based feature selection was implemented to extract AD-related anatomical features while excluding non-significant features. Then, seven knowledge-based ROIs were used to capture the local discriminative powers of selected voxels within areas that were sensitive to AD or mild cognitive impairment (MCI). The discriminative capability of the proposed feature was measured by its performance in differentiating AD or MCI from normal elderly controls (NC) using a support vector machine. The statistic-based feature selection, together with the knowledge-based masks, provided a promising solution for capturing anatomical features of the brain efficiently. For the analysis of clinical populations, which are inherently heterogeneous, this approach could stratify the large amount of data rapidly and could be combined with more detailed subsequent analyses based on non-linear transformation.

Most existing genome-wide association analyses are cross-sectional, utilizing only phenotypic data at a single time point, e.g. baseline. On the other hand, longitudinal studies, such as Alzheimer's Disease Neuroimaging Initiative (ADNI), collect phenotypic information at multiple time points. In this article, as a case study, we conducted both longitudinal and cross-sectional analyses of the ADNI data with several brain imaging (not clinical diagnosis) phenotypes, demonstrating the power gains of longitudinal analysis over cross-sectional analysis. Specifically, we scanned genome-wide single nucleotide polymorphisms (SNPs) with 56 brain-wide imaging phenotypes processed by FreeSurfer on 638 subjects. At the genome-wide significance level P < 1.8 x 10(9)) or a less stringent level (e.g. P < 10(7)), longitudinal analysis of the phenotypic data from the baseline to month 48 identified more SNP-phenotype associations than cross-sectional analysis of only the baseline data. In particular, at the genome-wide significance level, both SNP rs429358 in gene APOE and SNP rs2075650 in gene TOMM40 were confirmed to be associated with various imaging phenotypes in multiple regions of interests (ROIs) by both analyses, though longitudinal analysis detected more regional phenotypes associated with the two SNPs and indicated another significant SNP rs439401 in gene APOE. In light of the power advantage of longitudinal analysis, we advocate its use in current and future longitudinal neuroimaging studies.

Mild cognitive impairment (MCI) is a heterogeneous group and certain MCI subsets eventually convert to dementia. Cerebrospinal fluid (CSF) biomarkers are known to predict this conversion. We sought evidence for the differences in white matter connectivity between early amnestic MCI (EMCI) subgroups according to a CSF phosphorylated tau181p/amyloid beta1-42 ratio of 0.10. From the Alzheimer's Disease Neuroimaging Initiative database, 16 high-ratio, 25 low-ratio EMCI patients, and 20 normal controls with diffusion tensor images and CSF profiles were included. Compared to the high-ratio group, radial diffusivity significantly increased in both sides of the corpus callosum and the superior and inferior longitudinal fasciculus in the low-ratio group. In widespread white matter skeleton regions, the low-ratio group showed significantly increased mean, axial, and radial diffusivity compared to normal controls. However, the high-ratio group showed no differences when compared to the normal group. In conclusion, our study revealed that there were significant differences in white matter connectivity between EMCI subgroups according to CSF phosphorylated tau181p/amyloid beta1-42 ratios.

Changes in the levels of circulating proteins are associated with Alzheimer's disease (AD), whereas their pathogenic roles in AD are unclear. Here, we identified soluble ST2 (sST2), a decoy receptor of interleukin-33-ST2 signaling, as a new disease-causing factor in AD. Increased circulating sST2 level is associated with more severe pathological changes in female individuals with AD. Genome-wide association analysis and CRISPR-Cas9 genome editing identified rs1921622 , a genetic variant in an enhancer element of IL1RL1, which downregulates gene and protein levels of sST2. Mendelian randomization analysis using genetic variants, including rs1921622 , demonstrated that decreased sST2 levels lower AD risk and related endophenotypes in females carrying the Apolipoprotein E (APOE)-ε4 genotype; the association is stronger in Chinese than in European-descent populations. Human and mouse transcriptome and immunohistochemical studies showed that rs1921622 /sST2 regulates amyloid-beta (Aβ) pathology through the modulation of microglial activation and Aβ clearance. These findings demonstrate how sST2 level is modulated by a genetic variation and plays a disease-causing role in females with AD.

The failure of current strategies to provide an explanation for controversial findings on the pattern of pathophysiological changes in Alzheimer's Disease (AD) motivates the necessity to develop new integrative approaches based on multi-modal neuroimaging data that captures various aspects of disease pathology. Previous studies using [18F]fluorodeoxyglucose positron emission tomography (FDG-PET) and structural magnetic resonance imaging (sMRI) report controversial results about time-line, spatial extent and magnitude of glucose hypometabolism and atrophy in AD that depend on clinical and demographic characteristics of the studied populations. Here, we provide and validate at a group level a generative anatomical model of glucose hypo-metabolism and atrophy progression in AD based on FDG-PET and sMRI data of 80 patients and 79 healthy controls to describe expected age and symptom severity related changes in AD relative to a baseline provided by healthy aging. We demonstrate a high level of anatomical accuracy for both modalities yielding strongly age- and symptom-severity- dependant glucose hypometabolism in temporal, parietal and precuneal regions and a more extensive network of atrophy in hippocampal, temporal, parietal, occipital and posterior caudate regions. The model suggests greater and more consistent changes in FDG-PET compared to sMRI at earlier and the inversion of this pattern at more advanced AD stages. Our model describes, integrates and predicts characteristic patterns of AD related pathology, uncontaminated by normal age effects, derived from multi-modal data. It further provides an integrative explanation for findings suggesting a dissociation between early- and late-onset AD. The generative model offers a basis for further development of individualized biomarkers allowing accurate early diagnosis and treatment evaluation.

Longitudinal brain image analysis is critical for revealing subtle but complex structural and functional changes of brain during aging or in neurodevelopmental disease. However, even with the rapid increase of clinical research and trials, a software toolbox dedicated for longitudinal image analysis is still lacking publicly. To cater for this increasing need, we have developed a dedicated 4D Adult Brain Extraction and Analysis Toolbox (aBEAT) to provide robust and accurate analysis of the longitudinal adult brain MR images. Specially, a group of image processing tools were integrated into aBEAT, including 4D brain extraction, 4D tissue segmentation, and 4D brain labeling. First, a 4D deformable-surface-based brain extraction algorithm, which can deform serial brain surfaces simultaneously under temporal smoothness constraint, was developed for consistent brain extraction. Second, a level-sets-based 4D tissue segmentation algorithm that incorporates local intensity distribution, spatial cortical-thickness constraint, and temporal cortical-thickness consistency was also included in aBEAT for consistent brain tissue segmentation. Third, a longitudinal groupwise image registration framework was further integrated into aBEAT for consistent ROI labeling by simultaneously warping a pre-labeled brain atlas to the longitudinal brain images. The performance of aBEAT has been extensively evaluated on a large number of longitudinal MR T1 images which include normal and dementia subjects, achieving very promising results. A Linux-based standalone package of aBEAT is now freely available at http://www.nitrc.org/projects/abeat.

ObjectiveA positive family history (FH) is a risk factor for late-onset Alzheimer's disease (AD). Our aim was to examine the effects of FH on pathological and neuronal loss biomarkers across the cognitive spectrum.DesignCross-sectional analyses of data from a national biomarker study.SettingThe Alzheimer's Disease Neuroimaging Initiative national study.Patients257 subjects (ages 55-89), divided into cognitively normal (CN), mild cognitive impairment (MCI), and AD groups, with CSF and FH data.Outcome measuresCerebrospinal fluid (CSF) Aβ42, tau, and tau/Aβ42 ratio, MRI-measured hippocampal volumes.StatisticsUnivariate and multivariate analyses.ResultsIn MCI, CSF Aβ42 was lower (p = .005), t-tau was higher (p = 0.02) and t-tau/Aβ42 ratio was higher (p = 0.002) in FH+ than FH- subjects. A significant residual effect of FH on pathologic markers in MCI remained after adjusting for ApoE4 (p0.39) versus 21% of FH- controls (p = 0.03). The FH effect was not significant in AD subjects. Hippocampal and intracranial volumes did not differ between FH+ and FH- subjects in any group.ConclusionsA positive family history of late-onset AD is associated with a higher prevalence of an abnormal cerebral beta-amyloid and tau protein phenotype in MCI. The unexplained genetic heritability in family history is about the half the size of the ApoE4 effect. Longitudinal studies are warranted to more definitively examine this issue.

Different inflammatory and metabolic pathways have been associated with Alzheimeŕs disease (AD). However, only recently multi-analyte panels to study a large number of molecules in well characterized cohorts have been made available. These panels could help identify molecules that point to the affected pathways. We studied the relationship between a panel of plasma biomarkers (Human DiscoveryMAP) and presence of AD-like brain atrophy patterns defined by a previously published index (SPARE-AD) at baseline in subjects of the ADNI cohort. 818 subjects had MRI-derived SPARE-AD scores, of these subjects 69% had plasma biomarkers and 51% had CSF tau and Aβ measurements. Significant analyte-SPARE-AD and analytes correlations were studied in adjusted models. Plasma cortisol and chromogranin A showed a significant association that did not remain significant in the CSF signature adjusted model. Plasma macrophage inhibitory protein-1α and insulin-like growth factor binding protein 2 showed a significant association with brain atrophy in the adjusted model. Cortisol levels showed an inverse association with tests measuring processing speed. Our results indicate that stress and insulin responses and cytokines associated with recruitment of inflammatory cells in MCI-AD are associated with its characteristic AD-like brain atrophy pattern and correlate with clinical changes or CSF biomarkers.

There is accumulating evidence that neurotrophins, like brain-derived neurotrophic factor (BDNF), may impact aging and Alzheimer's Disease. However, traditional genetic association studies have not found a clear relationship between BDNF and AD. Our goal was to test whether BDNF single nucleotide polymorphisms (SNPs) impact Alzheimer's Disease-related brain imaging and cognitive markers of disease. We completed an imaging genetics study on 645 Alzheimer's Disease Neuroimaging Initiative participants (ND=175, MCI=316, AD=154) who had cognitive, brain imaging, and genetics data at baseline and a subset of those with brain imaging data at two years. Samples were genotyped using the Illumina Human610-Quad BeadChip. 13 SNPs in BDNF were identified in the dataset following quality control measures (rs6265(Val66Met), rs12273363, rs11030094, rs925946, rs1050187, rs2203877, rs11030104, rs11030108, rs10835211, rs7934165, rs908867, rs1491850, rs1157459). We analyzed a subgroup of 8 SNPs that were in low linkage disequilibrium with each other. Automated brain morphometric measures were available through ADNI investigators, and we analyzed baseline cognitive scores, hippocampal and whole brain volumes, and rates of hippocampal and whole brain atrophy and rates of change in the ADAS-Cog over one and two years. Three out of eight BDNF SNPs analyzed were significantly associated with measures of cognitive decline (rs1157659, rs11030094, rs11030108). No SNPs were significantly associated with baseline brain volume measures, however six SNPs were significantly associated with hippocampal and/or whole brain atrophy over two years (rs908867, rs11030094, rs6265, rs10501087, rs1157659, rs1491850). We also found an interaction between the BDNF Val66Met SNP and age with whole brain volume. Our imaging-genetics analysis in a large dataset suggests that while BDNF genetic variation is not specifically associated with a diagnosis of AD, it appears to play a role in AD-related brain neurodegeneration.

Multi-atlas segmentation has been widely used to segment various anatomical structures. The success of this technique partly relies on the selection of atlases that are best mapped to a new target image after registration. Recently, manifold learning has been proposed as a method for atlas selection. Each manifold learning technique seeks to optimize a unique objective function. Therefore, different techniques produce different embeddings even when applied to the same data set. Previous studies used a single technique in their method and gave no reason for the choice of the manifold learning technique employed nor the theoretical grounds for the choice of the manifold parameters. In this study, we compare side-by-side the results given by 3 manifold learning techniques (Isomap, Laplacian Eigenmaps and Locally Linear Embedding) on the same data set. We assess the ability of those 3 different techniques to select the best atlases to combine in the framework of multi-atlas segmentation. First, a leave-one-out experiment is used to optimize our method on a set of 110 manually segmented atlases of hippocampi and find the manifold learning technique and associated manifold parameters that give the best segmentation accuracy. Then, the optimal parameters are used to automatically segment 30 subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI). For our dataset, the selection of atlases with Locally Linear Embedding gives the best results. Our findings show that selection of atlases with manifold learning leads to segmentation accuracy close to or significantly higher than the state-of-the-art method and that accuracy can be increased by fine tuning the manifold learning process.

This paper aims to study the abnormal patterns of brain glucose metabolism co-variations in Alzheimer disease (AD) and Mild Cognitive Impairment (MCI) patients compared to Normal healthy controls (NC) using the Alzheimer Disease Neuroimaging Initiative (ADNI) database. The local cerebral metabolic rate for glucose (CMRgl) in a set of 90 structures belonging to the AAL atlas was obtained from Fluro-Deoxyglucose Positron Emission Tomography data in resting state. It is assumed that brain regions whose CMRgl values are significantly correlated are functionally associated; therefore, when metabolism is altered in a single region, the alteration will affect the metabolism of other brain areas with which it interrelates. The glucose metabolism network (represented by the matrix of the CMRgl co-variations among all pairs of structures) was studied using the graph theory framework. The highest concurrent fluctuations in CMRgl were basically identified between homologous cortical regions in all groups. Significant differences in CMRgl co-variations in AD and MCI groups as compared to NC were found. The AD and MCI patients showed aberrant patterns in comparison to NC subjects, as detected by global and local network properties (global and local efficiency, clustering index, and others). MCI network's attributes showed an intermediate position between NC and AD, corroborating it as a transitional stage from normal aging to Alzheimer disease. Our study is an attempt at exploring the complex association between glucose metabolism, CMRgl covariations and the attributes of the brain network organization in AD and MCI.

The primary constituents of plaques (Aβ42/Aβ40) and neurofibrillary tangles (tau and phosphorylated forms of tau [ptau]) are the current leading diagnostic and prognostic cerebrospinal fluid (CSF) biomarkers for AD. In this study, we performed deep sequencing of APP, PSEN1, PSEN2, GRN, APOE and MAPT genes in individuals with extreme CSF Aβ42, tau, or ptau levels. One known pathogenic mutation (PSEN1 p.A426P), four high-risk variants for AD (APOE p.L46P, MAPT p.A152T, PSEN2 p.R62H and p.R71W) and nine novel variants were identified. Surprisingly, a coding variant in PSEN1, p.E318G (rs17125721-G) exhibited a significant association with high CSF tau (p = 9.2 × 10(-4)) and ptau (p = 1.8 × 10(-3)) levels. The association of the p.E318G variant with Aβ deposition was observed in APOE-ε4 allele carriers. Furthermore, we found that in a large case-control series (n = 5,161) individuals who are APOE-ε4 carriers and carry the p.E318G variant are at a risk of developing AD (OR = 10.7, 95% CI = 4.7-24.6) that is similar to APOE-ε4 homozygous (OR = 9.9, 95% CI = 7.2.9-13.6), and double the risk for APOE-ε4 carriers that do not carry p.E318G (OR = 3.9, 95% CI = 3.4-4.4). The p.E318G variant is present in 5.3% (n = 30) of the families from a large clinical series of LOAD families (n = 565) and exhibited a higher frequency in familial LOAD (MAF = 2.5%) than in sporadic LOAD (MAF = 1.6%) (p = 0.02). Additionally, we found that in the presence of at least one APOE-ε4 allele, p.E318G is associated with more Aβ plaques and faster cognitive decline. We demonstrate that the effect of PSEN1, p.E318G on AD susceptibility is largely dependent on an interaction with APOE-ε4 and mediated by an increased burden of Aβ deposition.

BACKGROUND:Cerebrospinal fluid (CSF) α-synuclein is reduced in synucleinopathies, including dementia with Lewy bodies, and some studies have found increased CSF α-synuclein in Alzheimer's disease (AD). No study has explored effects of CSF α-synuclein on brain atrophy. Here we tested if baseline CSF α-synuclein affects brain atrophy rates and if these effects vary across brain regions, and across the cognitive spectrum from healthy elders (NL), to patients with mild cognitive impairment (MCI) and AD. METHODS:Baseline CSF α-synuclein measurements and longitudinal structural brain magnetic resonance imaging was performed in 74 NL, 118 MCI patients and 55 AD patients. Effects of baseline CSF α-synuclein on regional atrophy rates were tested in 1) four pre-hoc defined regions possibly associated with Lewy body and/or AD pathology (amygdala, caudate, hippocampus, brainstem), and 2) all available regions of interest. Differences across diagnoses were tested by assessing the interaction of CSF α-synuclein and diagnosis (testing NL versus MCI, and NL versus AD). RESULTS:The effects of CSF α-synuclein on longitudinal atrophy rates were not significant after correction for multiple comparisons. There were tendencies for effects in AD in caudate (higher atrophy rates in subjects with higher CSF α-synuclein, P=0.046) and brainstem (higher atrophy rates in subjects with lower CSF α-synuclein, P=0.063). CSF α-synuclein had significantly different effects on atrophy rates in NL and AD in brainstem (P=0.037) and caudate (P=0.006). DISCUSSION:With the possible exception of caudate and brainstem, the overall weak effects of CSF α-synuclein on atrophy rates in NL, MCI and AD argues against CSF α-synuclein as a biomarker related to longitudinal brain atrophy in these diagnostic groups. Any effects of CSF α-synuclein may be attenuated by possible simultaneous occurrence of AD-related neuronal injury and concomitant Lewy body pathology, which may elevate and reduce CSF α-synuclein levels, respectively.

To investigate whether there is a specific dose-dependent effect of the Apolipoprotein E (APOE) ε4 and ε2 alleles on hippocampal volume, across the cognitive spectrum, from normal aging to Alzheimer's Disease (AD).We analyzed MR and genetic data on 662 patients from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database-198 cognitively normal controls (CN), 321 mild-cognitive impairment (MCI) subjects, and 143 AD subjects-looking for dose-dependent effects of the ε4 and ε2 alleles on hippocampal volumes. Volumes were measured using a fully-automated algorithm applied to high resolution T1-weighted MR images. Statistical analysis consisted of a multivariate regression with repeated-measures model.There was a dose-dependent effect of the ε4 allele on hippocampal volume in AD (p = 0.04) and MCI (p = 0.02)-in both cases, each allele accounted for loss of >150 mm(3) (approximately 4%) of hippocampal volume below the mean volume for AD and MCI subjects with no such alleles (Cohen's d = -0.16 and -0.19 for AD and MCI, respectively). There was also a dose-dependent, main effect of the ε2 allele (p

Proteins, widely studied as potential biomarkers, play important roles in numerous physiological functions and diseases. Genetic variation may modulate corresponding protein levels and point to the role of these variants in disease pathophysiology. Effects of individual single nucleotide polymorphisms (SNPs) within a gene were analyzed for corresponding plasma protein levels using genome-wide association study (GWAS) genotype data and proteomic panel data with 132 quality-controlled analytes from 521 Caucasian participants in the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort. Linear regression analysis detected 112 significant (Bonferroni threshold p=2.44×10(-5)) associations between 27 analytes and 112 SNPs. 107 out of these 112 associations were tested in the Indiana Memory and Aging Study (IMAS) cohort for replication and 50 associations were replicated at uncorrected p

Alzheimer's disease (AD) is a well-known neurodegenerative disease that is associated with dramatic morphological abnormalities. The default mode network (DMN) is one of the most frequently studied resting-state networks. However, less is known about specific structural dependency or interactions among brain regions within the DMN in AD. In this study, we performed a Bayesian network (BN) analysis based on regional grey matter volumes to identify differences in structural interactions among core DMN regions in structural MRI data from 80 AD patients and 101 normal controls (NC). Compared to NC, the structural interactions between the medial prefrontal cortex (mPFC) and other brain regions, including the left inferior parietal cortex (IPC), the left inferior temporal cortex (ITC) and the right hippocampus (HP), were significantly reduced in the AD group. In addition, the AD group showed prominent increases in structural interactions from the left ITC to the left HP, the left HP to the right ITC, the right HP to the right ITC, and the right IPC to the posterior cingulate cortex (PCC). The BN models significantly distinguished AD patients from NC with 87.12% specificity and 81.25% sensitivity. We then used the derived BN models to examine the replicability and stability of AD-associated BN models in an independent dataset and the results indicated discriminability with 83.64% specificity and 80.49% sensitivity. The results revealed that the BN analysis was effective for characterising regional structure interactions and the AD-related BN models could be considered as valid and predictive structural brain biomarker models for AD. Therefore, our study can assist in further understanding the pathological mechanism of AD, based on the view of the structural network, and may provide new insights into classification and clinical application in the study of AD in the future.

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.

The goal of this work is to introduce new metrics to assess risk of Alzheimer's disease (AD) which we call AD Pattern Similarity (AD-PS) scores. These metrics are the conditional probabilities modeled by large-scale regularized logistic regression. The AD-PS scores derived from structural MRI and cognitive test data were tested across different situations using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) study. The scores were computed across groups of participants stratified by cognitive status, age and functional status. Cox proportional hazards regression was used to evaluate associations with the distribution of conversion times from mild cognitive impairment to AD. The performances of classifiers developed using data from different types of brain tissue were systematically characterized across cognitive status groups. We also explored the performance of anatomical and cognitive-anatomical composite scores generated by combining the outputs of classifiers developed using different types of data. In addition, we provide the AD-PS scores performance relative to other metrics used in the field including the Spatial Pattern of Abnormalities for Recognition of Early AD (SPARE-AD) index and total hippocampal volume for the variables examined.

Alzheimer's disease (AD), the most common form of dementia, shares many aspects of abnormal brain aging. We present a novel magnetic resonance imaging (MRI)-based biomarker that predicts the individual progression of mild cognitive impairment (MCI) to AD on the basis of pathological brain aging patterns. By employing kernel regression methods, the expression of normal brain-aging patterns forms the basis to estimate the brain age of a given new subject. If the estimated age is higher than the chronological age, a positive brain age gap estimation (BrainAGE) score indicates accelerated atrophy and is considered a risk factor for conversion to AD. Here, the BrainAGE framework was applied to predict the individual brain ages of 195 subjects with MCI at baseline, of which a total of 133 developed AD during 36 months of follow-up (corresponding to a pre-test probability of 68%). The ability of the BrainAGE framework to correctly identify MCI-converters was compared with the performance of commonly used cognitive scales, hippocampus volume, and state-of-the-art biomarkers derived from cerebrospinal fluid (CSF). With accuracy rates of up to 81%, BrainAGE outperformed all cognitive scales and CSF biomarkers in predicting conversion of MCI to AD within 3 years of follow-up. Each additional year in the BrainAGE score was associated with a 10% greater risk of developing AD (hazard rate: 1.10 [CI: 1.07-1.13]). Furthermore, the post-test probability was increased to 90% when using baseline BrainAGE scores to predict conversion to AD. The presented framework allows an accurate prediction even with multicenter data. Its fast and fully automated nature facilitates the integration into the clinical workflow. It can be exploited as a tool for screening as well as for monitoring treatment options.

Typically twin studies are used to investigate the aggregate effects of genetic and environmental influences on brain phenotypic measures. Although some phenotypic measures are highly heritable in twin studies, SNPs (single nucleotide polymorphisms) identified by genome-wide association studies (GWAS) account for only a small fraction of the heritability of these measures. We mapped the genetic variation (the proportion of phenotypic variance explained by variation among SNPs) of volumes of pre-defined regions across the whole brain, as explained by 512,905 SNPs genotyped on 747 adult participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI). We found that 85% of the variance of intracranial volume (ICV) (p = 0.04) was explained by considering all SNPs simultaneously, and after adjusting for ICV, total grey matter (GM) and white matter (WM) volumes had genetic variation estimates near zero (p = 0.5). We found varying estimates of genetic variation across 93 non-overlapping regions, with asymmetry in estimates between the left and right cerebral hemispheres. Several regions reported in previous studies to be related to Alzheimer's disease progression were estimated to have a large proportion of volumetric variance explained by the SNPs.

PURPOSE:To compare the accuracies of predicting AD conversion by using a decision support system (PredictAD tool) and current research criteria of prodromal AD as identified by combinations of episodic memory impairment of hippocampal type and visual assessment of medial temporal lobe atrophy (MTA) on MRI and CSF biomarkers. METHODS:Altogether 391 MCI cases (158 AD converters) were selected from the ADNI cohort. All the cases had baseline cognitive tests, MRI and/or CSF levels of Aβ1-42 and Tau. Using baseline data, the status of MCI patients (AD or MCI) three years later was predicted using current diagnostic research guidelines and the PredictAD software tool designed for supporting clinical diagnostics. The data used were 1) clinical criteria for episodic memory loss of the hippocampal type, 2) visual MTA, 3) positive CSF markers, 4) their combinations, and 5) when the PredictAD tool was applied, automatically computed MRI measures were used instead of the visual MTA results. The accuracies of diagnosis were evaluated with the diagnosis made 3 years later. RESULTS:The PredictAD tool achieved the overall accuracy of 72% (sensitivity 73%, specificity 71%) in predicting the AD diagnosis. The corresponding number for a clinician's prediction with the assistance of the PredictAD tool was 71% (sensitivity 75%, specificity 68%). Diagnosis with the PredictAD tool was significantly better than diagnosis by biomarkers alone or the combinations of clinical diagnosis of hippocampal pattern for the memory loss and biomarkers (p≤0.037). CONCLUSION:With the assistance of PredictAD tool, the clinician can predict AD conversion more accurately than the current diagnostic criteria.

The purpose of this project was to evaluate white matter degeneration and its impact on hippocampal structural connectivity in patients with amnestic mild cognitive impairment, non-amnestic mild cognitive impairment and Alzheimer's disease. We estimated white matter fractional anisotropy, mean diffusivity and hippocampal structural connectivity in two independent cohorts. The ADNI cohort included 108 subjects [25 cognitively normal, 21 amnestic mild cognitive impairment, 47 non-amnestic mild cognitive impairment and 15 Alzheimer's disease]. A second cohort included 34 subjects [15 cognitively normal and 19 amnestic mild cognitive impairment] recruited in Montreal. All subjects underwent clinical and neuropsychological assessment in addition to diffusion and T1 MRI. Individual fractional anisotropy and mean diffusivity maps were generated using FSL-DTIfit. In addition, hippocampal structural connectivity maps expressing the probability of connectivity between the hippocampus and cortex were generated using a pipeline based on FSL-probtrackX. Voxel-based group comparison statistics of fractional anisotropy, mean diffusivity and hippocampal structural connectivity were estimated using Tract-Based Spatial Statistics. The proportion of abnormal to total white matter volume was estimated using the total volume of the white matter skeleton. We found that in both cohorts, amnestic mild cognitive impairment patients had 27-29% white matter volume showing higher mean diffusivity but no significant fractional anisotropy abnormalities. No fractional anisotropy or mean diffusivity differences were observed between non-amnestic mild cognitive impairment patients and cognitively normal subjects. Alzheimer's disease patients had 66.3% of normalized white matter volume with increased mean diffusivity and 54.3% of the white matter had reduced fractional anisotropy. Reduced structural connectivity was found in the hippocampal connections to temporal, inferior parietal, posterior cingulate and frontal regions only in the Alzheimer's group. The severity of white matter degeneration appears to be higher in advanced clinical stages, supporting the construct that these abnormalities are part of the pathophysiological processes of Alzheimer's disease.

Various studies have suggested that the mitochondrial genome plays a role in late-onset Alzheimer's disease, although results are mixed. We used an endophenotype-based approach to further characterize mitochondrial genetic variation and its relationship to risk markers for Alzheimer's disease. We analyzed longitudinal data from non-demented, mild cognitive impairment, and late-onset Alzheimer's disease participants in the Alzheimer's Disease Neuroimaging Initiative with genetic, brain imaging, and behavioral data. We assessed the relationship of structural MRI and cognitive biomarkers with mitochondrial genome variation using TreeScanning, a haplotype-based approach that concentrates statistical power by analyzing evolutionarily meaningful groups (or clades) of haplotypes together for association with a phenotype. Four clades were associated with three different endophenotypes: whole brain volume, percent change in temporal pole thickness, and left hippocampal atrophy over two years. This is the first study of its kind to identify mitochondrial variation associated with brain imaging endophenotypes of Alzheimer's disease. Our results provide additional evidence that the mitochondrial genome plays a role in risk for Alzheimer's disease.

BACKGROUNDWe introduced a hypometabolic convergence index (HCI) to characterize in a single measurement the extent to which a person's fluorodeoxyglucose positron emission tomogram (FDG PET) corresponds to that in Alzheimer's disease (AD). Apolipoprotein E ε4 (APOE ε4) gene dose is associated with three levels of risk for late-onset AD. We explored the association between gene dose and HCI in cognitively normal ε4 homozygotes, heterozygotes, and non-carriers.METHODSAn algorithm was used to characterize and compare AD-related HCIs in cognitively normal individuals, including 36 ε4 homozygotes, 46 heterozygotes, and 78 non-carriers.RESULTSThese three groups differed significantly in their HCIs (ANOVA, p = 0.004), and there was a significant association between HCIs and gene dose (linear trend, p = 0.001).CONCLUSIONSThe HCI is associated with three levels of genetic risk for late-onset AD. This supports the possibility of using a single FDG PET measurement to help in the preclinical detection and tracking of AD.

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.

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.

Copy number variations (CNVs) are genomic regions that have added (duplications) or deleted (deletions) genetic material. They may overlap genes affecting their function and have been shown to be associated with disease. We previously investigated the role of CNVs in late-onset Alzheimer's disease (AD) and mild cognitive impairment using Alzheimer's Disease Neuroimaging Initiative (ADNI) and National Institute of Aging-Late Onset AD/National Cell Repository for AD (NIA-LOAD/NCRAD) Family Study participants, and identified a number of genes overlapped by CNV calls. To confirm the findings and identify other potential candidate regions, we analyzed array data from a unique cohort of 1617 Caucasian participants (1022 AD cases and 595 controls) who were clinically characterized and whose diagnosis was neuropathologically verified. All DNA samples were extracted from brain tissue. CNV calls were generated and subjected to quality control (QC). 728 cases and 438 controls who passed all QC measures were included in case/control association analyses including candidate gene and genome-wide approaches. Rates of deletions and duplications did not significantly differ between cases and controls. Case-control association identified a number of previously reported regions (CHRFAM7A, RELN and DOPEY2) as well as a new gene (HLA-DRA). Meta-analysis of CHRFAM7A indicated a significant association of the gene with AD and/or MCI risk (P = 0.006, odds ratio = 3.986 (95% confidence interval 1.490-10.667)). A novel APP gene duplication was observed in one case sample. Further investigation of the identified genes in independent and larger samples is warranted.

The tau and amyloid pathobiological processes underlying Alzheimer disease (AD) progresses slowly over periods of decades before clinical manifestation as mild cognitive impairment (MCI), then more rapidly to dementia, and eventually to end-stage organ failure. The failure of clinical trials of candidate disease modifying therapies to slow disease progression in patients already diagnosed with early AD has led to increased interest in exploring the possibility of early intervention and prevention trials, targeting MCI and cognitively healthy (HC) populations. Here, we stratify MCI individuals based on cerebrospinal fluid (CSF) biomarkers and structural atrophy risk factors for the disease. We also stratify HC individuals into risk groups on the basis of CSF biomarkers for the two hallmark AD pathologies. Results show that the broad category of MCI can be decomposed into subsets of individuals with significantly different average regional atrophy rates. By thus selectively identifying individuals, combinations of these biomarkers and risk factors could enable significant reductions in sample size requirements for clinical trials of investigational AD-modifying therapies, and provide stratification mechanisms to more finely assess response to therapy. Power is sufficiently high that detecting efficacy in MCI cohorts should not be a limiting factor in AD therapeutics research. In contrast, we show that sample size estimates for clinical trials aimed at the preclinical stage of the disorder (HCs with evidence of AD pathology) are prohibitively large. Longer natural history studies are needed to inform design of trials aimed at the presymptomatic stage.

BACKGROUND:Machine learning neuroimaging researchers have often relied on regularization techniques when classifying MRI images. Although these were originally introduced to deal with "ill-posed" problems it is rare to find studies that evaluate the ill-posedness of MRI image classification problems. In addition, to avoid the effects of the "curse of dimensionality" very often dimension reduction is applied to the data. METHODOLOGY:Baseline structural MRI data from cognitively normal and Alzheimer's disease (AD) patients from the AD Neuroimaging Initiative database were used in this study. We evaluated here the ill-posedness of this classification problem across different dimensions and sample sizes and its relationship to the performance of regularized logistic regression (RLR), linear support vector machine (SVM) and linear regression classifier (LRC). In addition, these methods were compared with their principal components space counterparts. PRINCIPAL FINDINGS:In voxel space the prediction performance of all methods increased as sample sizes increased. They were not only relatively robust to the increase of dimension, but they often showed improvements in accuracy. We linked this behavior to improvements in conditioning of the linear kernels matrices. In general the RLR and SVM performed similarly. Surprisingly, the LRC was often very competitive when the linear kernel matrices were best conditioned. Finally, when comparing these methods in voxel and principal component spaces, we did not find large differences in prediction performance. CONCLUSIONS AND SIGNIFICANCE:We analyzed the problem of classifying AD MRI images from the perspective of linear ill-posed problems. We demonstrate empirically the impact of the linear kernel matrix conditioning on different classifiers' performance. This dependence is characterized across sample sizes and dimensions. In this context we also show that increased dimensionality does not necessarily degrade performance of machine learning methods. In general, this depends on the nature of the problem and the type of machine learning method.

The hypothetical model of dynamic biomarkers for Alzheimer's disease (AD) describes high amyloid deposition and hypometabolism at the mild cognitive impairment (MCI) stage. However, it remains unknown whether brain amyloidosis and hypometabolism follow the same trajectories in MCI individuals. We used the concept of early MCI (EMCI) and late MCI (LMCI) as defined by the Alzheimer's disease Neuroimaging Initiative (ADNI)-Go in order to compare the biomarker profile between EMCI and LMCI.To examine the global and voxel-based neocortical amyloid burden and metabolism among individuals who are cognitively normal (CN), as well as those with EMCI, LMCI and mild AD.In the present study, 354 participants, including CN (n = 109), EMCI (n = 157), LMCI (n = 39) and AD (n = 49), were enrolled between September 2009 and November 2011 through ADNI-GO and ADNI-2. Brain amyloid load and metabolism were estimated using [(18)F]AV45 and [(18)F]fluorodeoxyglucose ([(18)F]FDG) PET, respectively. Uptake ratio images of [(18)F]AV45 and [(18)F]FDG were calculated by dividing the summed PET image by the median counts of the grey matter of the cerebellum and pons, respectively. Group differences of global [(18)F]AV45 and [(18)F]FDG were analyzed using ANOVA, while the voxel-based group differences were estimated using statistic parametric mapping (SPM).EMCI patients showed higher global [(18)F]AV45 retention compared to CN and lower uptake compared to LMCI. SPM detected higher [(18)F]AV45 uptake in EMCI compared to CN in the precuneus, posterior cingulate, medial and dorsal lateral prefrontal cortices, bilaterally. EMCI showed lower [(18)F]AV45 retention than LMCI in the superior temporal, inferior parietal, as well as dorsal lateral prefrontal cortices, bilaterally. Regarding to the global [(18)F]FDG, EMCI patients showed no significant difference from CN and a higher uptake ratio compared to LMCI. At the voxel level, EMCI showed higher metabolism in precuneus, hippocampus, entorhinal and inferior parietal cortices, as compared to LMCI.The present results indicate that brain metabolism remains normal despite the presence of significant amyloid accumulation in EMCI. These results suggest a role for anti-amyloid interventions in EMCI aiming to delay or halt the deposition of amyloid and related metabolism impairment.

Diagnosis of Alzheimer's disease is based on the results of neuropsychological tests and available supporting biomarkers such as the results of imaging studies. The results of the tests and the values of biomarkers are dependent on the nuisance features, such as age and gender. In order to improve diagnostic power, the effects of the nuisance features have to be removed from the data. In this paper, four types of interactions between classification features and nuisance features were identified. Three methods were tested to remove these interactions from the classification data. In stratified analysis, a homogeneous subgroup was generated from a training set. Data correction method utilized linear regression model to remove the effects of nuisance features from data. The third method was a combination of these two methods. The methods were tested using all the baseline data from the Alzheimer's Disease Neuroimaging Initiative database in two classification studies: classifying control subjects from Alzheimer's disease patients and discriminating stable and progressive mild cognitive impairment subjects. The results show that both stratified analysis and data correction are able to statistically significantly improve the classification accuracy of several neuropsychological tests and imaging biomarkers. The improvements were especially large for the classification of stable and progressive mild cognitive impairment subjects, where the best improvements observed were 6% units. The data correction method gave better results for imaging biomarkers, whereas stratified analysis worked well with the neuropsychological tests. In conclusion, the study shows that the excess variability caused by nuisance features should be removed from the data to improve the classification accuracy, and therefore, the reliability of diagnosis making.

BACKGROUND:Recent studies showed that polymorphisms in the Fat and Obesity-Associated (FTO) gene have robust effects on obesity, obesity-related traits and endophenotypes associated with Alzheimer's disease (AD). METHODS:We used 1,877 Caucasian cases and controls from the NIA-LOAD study and 1,093 Caribbean Hispanics to further explore the association of FTO with AD. Using logistic regression, we assessed 42 SNPs in introns 1 and 2, the region previously reported to be associated with AD endophenotypes, which had been derived by genome-wide screenings. In addition, we performed gene expression analyses of neuropathologically confirmed AD cases and controls of two independent datasets (19 AD cases, 10 controls; 176 AD cases, 188 controls) using within- and between-group factors ANOVA of log(10) transformed rank invariant normalized expression data. RESULTS:In the NIALOAD study, one SNP was significantly associated with AD and three additional markers were close to significance (rs6499640, rs10852521, rs16945088, rs8044769, FDR p-value: 0.050.9) with the previously reported SNPs. In the Caribbean Hispanic dataset, we identified three SNPs (rs17219084, rs11075996, rs11075997, FDR p-value: 0.009

Age is the strongest risk factor for sporadic Alzheimer disease (AD), yet the effects of age on rates of clinical decline and brain atrophy in AD have been largely unexplored. Here, we examined longitudinal rates of change as a function of baseline age for measures of clinical decline and structural MRI-based regional brain atrophy, in cohorts of AD, mild cognitive impairment (MCI), and cognitively healthy (HC) individuals aged 65 to 90 years (total n = 723). The effect of age was modeled using mixed effects linear regression. There was pronounced reduction in rates of clinical decline and atrophy with age for AD and MCI individuals, whereas HCs showed increased rates of clinical decline and atrophy with age. This resulted in convergence in rates of change for HCs and patients with advancing age for several measures. Baseline cerebrospinal fluid densities of AD-relevant proteins, Aβ(1-42), tau, and phospho-tau(181p) (ptau), showed a similar pattern of convergence with advanced age across cohorts, particularly for ptau. In contrast, baseline clinical measures did not differ by age, indicating uniformity of clinical severity at baseline. These results imply that the phenotypic expression of AD is relatively mild in individuals older than approximately 85 years, and this may affect the ability to distinguish AD from normal aging in the very old. Our findings show that inclusion of older individuals in clinical trials will substantially reduce the power to detect disease-modifying therapeutic effects, leading to dramatic increases in required clinical trial sample sizes with age of study sample.

BACKGROUND:Recent Alzheimer's disease (AD) research has focused on finding biomarkers to identify disease at the pre-clinical stage of mild cognitive impairment (MCI), allowing treatment to be initiated before irreversible damage occurs. Many studies have examined brain imaging or cerebrospinal fluid but there is also growing interest in blood biomarkers. The Alzheimer's Disease Neuroimaging Initiative (ADNI) has generated data on 190 plasma analytes in 566 individuals with MCI, AD or normal cognition. We conducted independent analyses of this dataset to identify plasma protein signatures predicting pre-clinical AD. METHODS AND FINDINGS:We focused on identifying signatures that discriminate cognitively normal controls (n = 54) from individuals with MCI who subsequently progress to AD (n = 163). Based on p value, apolipoprotein E (APOE) showed the strongest difference between these groups (p = 2.3 × 10(-13)). We applied a multivariate approach based on combinatorial optimization ((α,β)-k Feature Set Selection), which retains information about individual participants and maintains the context of interrelationships between different analytes, to identify the optimal set of analytes (signature) to discriminate these two groups. We identified 11-analyte signatures achieving values of sensitivity and specificity between 65% and 86% for both MCI and AD groups, depending on whether APOE was included and other factors. Classification accuracy was improved by considering "meta-features," representing the difference in relative abundance of two analytes, with an 8-meta-feature signature consistently achieving sensitivity and specificity both over 85%. Generating signatures based on longitudinal rather than cross-sectional data further improved classification accuracy, returning sensitivities and specificities of approximately 90%. CONCLUSIONS:Applying these novel analysis approaches to the powerful and well-characterized ADNI dataset has identified sets of plasma biomarkers for pre-clinical AD. While studies of independent test sets are required to validate the signatures, these analyses provide a starting point for developing a cost-effective and minimally invasive test capable of diagnosing AD in its pre-clinical stages.

One of the most significant evolutionary changes underlying the highly developed cognitive abilities of humans is the greatly enlarged brain volume. In addition to being far greater than in most other species, the volume of the human brain exhibits extensive variation and distinct sexual dimorphism in the general population. However, little is known about the genetic mechanisms underlying normal variation as well as the observed sex difference in human brain volume. Here we show that interleukin-3 (IL3) is strongly associated with brain volume variation in four genetically divergent populations. We identified a sequence polymorphism (rs31480) in the IL3 promoter which alters the expression of IL3 by affecting the binding affinity of transcription factor SP1. Further analysis indicated that IL3 and its receptors are continuously expressed in the developing mouse brain, reaching highest levels at postnatal day 1-4. Furthermore, we found IL3 receptor alpha (IL3RA) was mainly expressed in neural progenitors and neurons, and IL3 could promote proliferation and survival of the neural progenitors. The expression level of IL3 thus played pivotal roles in the expansion and maintenance of the neural progenitor pool and the number of surviving neurons. Moreover, we found that IL3 activated both estrogen receptors, but estrogen didn't directly regulate the expression of IL3. Our results demonstrate that genetic variation in the IL3 promoter regulates human brain volume and reveals novel roles of IL3 in regulating brain development.

Recent genome-wide association studies of Alzheimer's disease (AD) have identified variants in BIN1, CLU, CR1 and PICALM that show replicable association with risk for disease. We have thoroughly sampled common variation in these genes, genotyping 355 variants in over 600 individuals for whom measurements of two AD biomarkers, cerebrospinal fluid (CSF) 42 amino acid amyloid beta fragments (Aβ(42)) and tau phosphorylated at threonine 181 (ptau(181)), have been obtained. Association analyses were performed to determine whether variants in BIN1, CLU, CR1 or PICALM are associated with changes in the CSF levels of these biomarkers. Despite adequate power to detect effects as small as a 1.05 fold difference, we have failed to detect evidence for association between SNPs in these genes and CSF Aβ(42) or ptau(181) levels in our sample. Our results suggest that these variants do not affect risk via a mechanism that results in a strong additive effect on CSF levels of Aβ(42) or ptau(181).

Recent GWAS studies focused on uncovering novel genetic loci related to AD have revealed associations with variants near CLU, CR1, PICALM and BIN1. In this study, we conducted a genome-wide association study in an independent set of 1034 cases and 1186 controls using the Illumina genotyping platforms. By coupling our data with available GWAS datasets from the ADNI and GenADA, we replicated the original associations in both PICALM (rs3851179) and CR1 (rs3818361). The PICALM variant seems to be non-significant after we adjusted for APOE e4 status. We further tested our top markers in 751 independent cases and 751 matched controls. Besides the markers close to the APOE locus, a marker (rs12989701) upstream of BIN1 locus was replicated and the combined analysis reached genome-wide significance level (p = 5E-08). We combined our data with the published Harold et al. study and meta-analysis with all available 6521 cases and 10360 controls at the BIN1 locus revealed two significant variants (rs12989701, p = 1.32E-10 and rs744373, p = 3.16E-10) in limited linkage disequilibrium (r² = 0.05) with each other. The independent contribution of both SNPs was supported by haplotype conditional analysis. We also conducted multivariate analysis in canonical pathways and identified a consistent signal in the downstream pathways targeted by Gleevec (P = 0.004 in Pfizer; P = 0.028 in ADNI and P = 0.04 in GenADA). We further tested variants in CLU, PICALM, BIN1 and CR1 for association with disease progression in 597 AD patients where longitudinal cognitive measures are sufficient. Both the PICALM and CLU variants showed nominal significant association with cognitive decline as measured by change in Clinical Dementia Rating-sum of boxes (CDR-SB) score from the baseline but did not pass multiple-test correction. Future experiments will help us better understand potential roles of these genetic loci in AD pathology.

ContextThere is no rapid and cost effective tool that can be implemented as a front-line screening tool for Alzheimer's disease (AD) at the population level.ObjectiveTo generate and cross-validate a blood-based screener for AD that yields acceptable accuracy across both serum and plasma.Design, setting, participantsAnalysis of serum biomarker proteins were conducted on 197 Alzheimer's disease (AD) participants and 199 control participants from the Texas Alzheimer's Research Consortium (TARC) with further analysis conducted on plasma proteins from 112 AD and 52 control participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI). The full algorithm was derived from a biomarker risk score, clinical lab (glucose, triglycerides, total cholesterol, homocysteine), and demographic (age, gender, education, APOE*E4 status) data.Major outcome measuresAlzheimer's disease.Results11 proteins met our criteria and were utilized for the biomarker risk score. The random forest (RF) biomarker risk score from the TARC serum samples (training set) yielded adequate accuracy in the ADNI plasma sample (training set) (AUC = 0.70, sensitivity (SN) = 0.54 and specificity (SP) = 0.78), which was below that obtained from ADNI cerebral spinal fluid (CSF) analyses (t-tau/Aβ ratio AUC = 0.92). However, the full algorithm yielded excellent accuracy (AUC = 0.88, SN = 0.75, and SP = 0.91). The likelihood ratio of having AD based on a positive test finding (LR+) = 7.03 (SE = 1.17; 95% CI = 4.49-14.47), the likelihood ratio of not having AD based on the algorithm (LR-) = 3.55 (SE = 1.15; 2.22-5.71), and the odds ratio of AD were calculated in the ADNI cohort (OR) = 28.70 (1.55; 95% CI = 11.86-69.47).ConclusionsIt is possible to create a blood-based screening algorithm that works across both serum and plasma that provides a comparable screening accuracy to that obtained from CSF analyses.

The role of structural brain magnetic resonance imaging (MRI) is becoming more and more emphasized in the early diagnostics of Alzheimer's disease (AD). This study aimed to assess the improvement in classification accuracy that can be achieved by combining features from different structural MRI analysis techniques. Automatically estimated MR features used are hippocampal volume, tensor-based morphometry, cortical thickness and a novel technique based on manifold learning. Baseline MRIs acquired from all 834 subjects (231 healthy controls (HC), 238 stable mild cognitive impairment (S-MCI), 167 MCI to AD progressors (P-MCI), 198 AD) from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database were used for evaluation. We compared the classification accuracy achieved with linear discriminant analysis (LDA) and support vector machines (SVM). The best results achieved with individual features are 90% sensitivity and 84% specificity (HC/AD classification), 64%/66% (S-MCI/P-MCI) and 82%/76% (HC/P-MCI) with the LDA classifier. The combination of all features improved these results to 93% sensitivity and 85% specificity (HC/AD), 67%/69% (S-MCI/P-MCI) and 86%/82% (HC/P-MCI). Compared with previously published results in the ADNI database using individual MR-based features, the presented results show that a comprehensive analysis of MRI images combining multiple features improves classification accuracy and predictive power in detecting early AD. The most stable and reliable classification was achieved when combining all available features.

Owning to its clinical accessibility, T1-weighted MRI (Magnetic Resonance Imaging) has been extensively studied in the past decades for prediction of Alzheimer's disease (AD) and mild cognitive impairment (MCI). The volumes of gray matter (GM), white matter (WM) and cerebrospinal fluid (CSF) are the most commonly used measurements, resulting in many successful applications. It has been widely observed that disease-induced structural changes may not occur at isolated spots, but in several inter-related regions. Therefore, for better characterization of brain pathology, we propose in this paper a means to extract inter-regional correlation based features from local volumetric measurements. Specifically, our approach involves constructing an anatomical brain network for each subject, with each node representing a Region of Interest (ROI) and each edge representing Pearson correlation of tissue volumetric measurements between ROI pairs. As second order volumetric measurements, network features are more descriptive but also more sensitive to noise. To overcome this limitation, a hierarchy of ROIs is used to suppress noise at different scales. Pairwise interactions are considered not only for ROIs with the same scale in the same layer of the hierarchy, but also for ROIs across different scales in different layers. To address the high dimensionality problem resulting from the large number of network features, a supervised dimensionality reduction method is further employed to embed a selected subset of features into a low dimensional feature space, while at the same time preserving discriminative information. We demonstrate with experimental results the efficacy of this embedding strategy in comparison with some other commonly used approaches. In addition, although the proposed method can be easily generalized to incorporate other metrics of regional similarities, the benefits of using Pearson correlation in our application are reinforced by the experimental results. Without requiring new sources of information, our proposed approach improves the accuracy of MCI prediction from 80.83% (of conventional volumetric features) to 84.35% (of hierarchical network features), evaluated using data sets randomly drawn from the ADNI (Alzheimer's Disease Neuroimaging Initiative) dataset.