We aimed to improve on the single-atlas ventricular segmentation method of (Carmichael, O.T., Thompson, P.M., Dutton, R.A., Lu, A., Lee, S.E., Lee, J.Y., Kuller, L.H., Lopez, O.L., Aizenstein, H.J., Meltzer, C.C., Liu, Y., Toga, A.W., Becker, J.T., 2006. Mapping ventricular changes related to dementia and mild cognitive impairment in a large community-based cohort. IEEE ISBI. 315-318) by using multi-atlas segmentation, which has been shown to lead to more accurate segmentations (Chou, Y., Leporé, N., de Zubicaray, G., Carmichael, O., Becker, J., Toga, A., Thompson, P., 2008. Automated ventricular mapping with multi-atlas fluid image alignment reveals genetic effects in Alzheimer's disease, NeuroImage 40(2): 615-630); with this method, we calculated minimal numbers of subjects needed to detect correlations between clinical scores and ventricular maps. We also assessed correlations between emerging CSF biomarkers of Alzheimer's disease pathology and localizable deficits in the brain, in 80 AD, 80 mild cognitive impairment (MCI), and 80 healthy controls from the Alzheimer's Disease Neuroimaging Initiative. Six expertly segmented images and their embedded parametric mesh surfaces were fluidly registered to each brain; segmentations were averaged within subjects to reduce errors. Surface-based statistical maps revealed powerful correlations between surface morphology and 4 variables: (1) diagnosis, (2) depression severity, (3) cognitive function at baseline, and (4) future cognitive decline over the following year. Cognitive function was assessed using the mini-mental state exam (MMSE), global and sum-of-boxes clinical dementia rating (CDR) scores, at baseline and 1-year follow-up. Lower CSF Abeta(1-42) protein levels, a biomarker of AD pathology assessed in 138 of the 240 subjects, were correlated with lateral ventricular expansion. Using false discovery rate (FDR) methods, 40 and 120 subjects, respectively, were needed to discriminate AD and MCI from normal groups. 120 subjects were required to detect correlations between ventricular enlargement and MMSE, global CDR, sum-of-boxes CDR and clinical depression scores. Ventricular expansion maps correlate with pathological and cognitive measures in AD, and may be useful in future imaging-based clinical trials.

Rates of brain atrophy derived from serial magnetic resonance (MR) studies may be used to assess therapies for Alzheimer's disease (AD). These measures may be confounded by changes in scanner voxel sizes. For this reason, the Alzheimer's Disease Neuroimaging Initiative (ADNI) included the imaging of a geometric phantom with every scan. This study compares voxel scaling correction using a phantom with correction using a 9 degrees of freedom (9DOF) registration algorithm. We took 129 pairs of baseline and 1-year repeat scans, and calculated the volume scaling correction, previously measured using the phantom. We used the registration algorithm to quantify any residual scaling errors, and found the algorithm to be unbiased, with no significant (p=0.97) difference between control (n=79) and AD subjects (n=50), but with a mean (SD) absolute volume change of 0.20 (0.20) % due to linear scalings. 9DOF registration was shown to be comparable to geometric phantom correction in terms of the effect on atrophy measurement and unbiased with respect to disease status. These results suggest that the additional expense and logistic effort of scanning a phantom with every patient scan can be avoided by registration-based scaling correction. Furthermore, based upon the atrophy rates in the AD subjects in this study, sample size requirements would be approximately 10-12% lower with (either) correction for voxel scaling than if no correction was used.

The ability to pool data from multiple MRI scanners is becoming increasingly important with the influx in multi-site research studies. Fast spin echo (FSE) dual spin echo sequences are often chosen for such studies based principally on their short acquisition time and the clinically useful contrasts they provide for assessing gross pathology. The practicality of measuring FSE-T2 relaxation properties has rarely been assessed. Here, FSE-T2 relaxation properties are examined across the three main scanner vendors (General Electric (GE), Philips, and Siemens). The American College of Radiology (ACR) phantom was scanned on four 1.5T platforms (two GE, one Philips, and one Siemens) to determine if the dual echo pulse sequence is susceptible to vendor-based variance. In addition, data from 85 subjects spanning the spectrum of normal aging, mild cognitive impairment (MCI), and Alzheimer's disease (AD) was obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) to affirm the presence of any phantom based between vendor variance and determine the relationship between this variance and disease. FSE-T2 relaxation properties, including peak FSE-T2 and histogram width, were calculated for each phantom and human subject. Direct correspondence was found between the phantom and human subject data. Peak FSE-T2 of Siemens scanners was consistently at least 20ms prolonged compared to GE and Philips. Siemens scanners showed broader FSE-T2 histograms than the other scanners. Greater variance was observed across GE scanners than either Philips or Siemens. FSE-T2 differences were much greater with scanner vendor than between diagnostic groups, as no significant changes in peak FSE-T2 or histogram width between normal aged, MCI, and AD subject groups were observed. These results indicate that whole brain histogram measures are not sensitive enough to detect FSE-T2 changes between normal aging, MCI, and AD and that FSE-T2 is highly variable across scanner vendors.

Imaging traits provide a powerful and biologically relevant substrate to examine the influence of genetics on the brain. Interest in genome-wide, brain-wide search for influential genetic variants is growing, but has mainly focused on univariate, SNP-based association tests. Moving to gene-based multivariate statistics, we can test the combined effect of multiple genetic variants in a single test statistic. Multivariate models can reduce the number of statistical tests in gene-wide or genome-wide scans and may discover gene effects undetectable with SNP-based methods. Here we present a gene-based method for associating the joint effect of single nucleotide polymorphisms (SNPs) in 18,044 genes across 31,662 voxels of the whole brain in 731 elderly subjects (mean age: 75.56±6.82SD years; 430 males) from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Structural MRI scans were analyzed using tensor-based morphometry (TBM) to compute 3D maps of regional brain volume differences compared to an average template image based on healthy elderly subjects. Using the voxel-level volume difference values as the phenotype, we selected the most significantly associated gene (out of 18,044) at each voxel across the brain. No genes identified were significant after correction for multiple comparisons, but several known candidates were re-identified, as were other genes highly relevant to brain function. GAB2, which has been previously associated with late-onset AD, was identified as the top gene in this study, suggesting the validity of the approach. This multivariate, gene-based voxelwise association study offers a novel framework to detect genetic influences on the brain.

This article introduces a hypometabolic convergence index (HCI) for the assessment of Alzheimer's disease (AD); compares it to other biological, cognitive and clinical measures; and demonstrates its promise to predict clinical decline in mild cognitive impairment (MCI) patients using data from the AD Neuroimaging Initiative (ADNI). The HCI is intended to reflect in a single measurement the extent to which the pattern and magnitude of cerebral hypometabolism in an individual's fluorodeoxyglucose positron emission tomography (FDG-PET) image correspond to that in probable AD patients, and is generated using a fully automated voxel-based image-analysis algorithm. HCIs, magnetic resonance imaging (MRI) hippocampal volume measurements, cerebrospinal fluid (CSF) assays, memory test scores, and clinical ratings were compared in 47 probable AD patients, 21 MCI patients who converted to probable AD within the next 18months, 76 MCI patients who did not, and 47 normal controls (NCs) in terms of their ability to characterize clinical disease severity and predict conversion rates from MCI to probable AD. HCIs were significantly different in the probable AD, MCI converter, MCI stable and NC groups (p=9e-17) and correlated with clinical disease severity. Using retrospectively characterized threshold criteria, MCI patients with either higher HCIs or smaller hippocampal volumes had the highest hazard ratios (HRs) for 18-month progression to probable AD (7.38 and 6.34, respectively), and those with both had an even higher HR (36.72). In conclusion, the HCI, alone or in combination with certain other biomarker measurements, has the potential to help characterize AD and predict subsequent rates of clinical decline. More generally, our conversion index strategy could be applied to a range of imaging modalities and voxel-based image-analysis algorithms.

We introduce a mass-univariate framework for the analysis of whole-brain structural trajectories using longitudinal Voxel-Based Morphometry data and Bayesian inference. Our approach to developmental and aging longitudinal studies characterizes heterogeneous structural growth/decline between and within groups. In particular, we propose a probabilistic generative model that parameterizes individual and ensemble average changes in brain structure using linear mixed-effects models of age and subject-specific covariates. Model inversion uses Expectation Maximization (EM), while voxelwise (empirical) priors on the size of individual differences are estimated from the data. Bayesian inference on individual and group trajectories is realized using Posterior Probability Maps (PPM). In addition to parameter inference, the framework affords comparisons of models with varying combinations of model order for fixed and random effects using model evidence. We validate the model in simulations and real MRI data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) project. We further demonstrate how subject specific characteristics contribute to individual differences in longitudinal volume changes in healthy subjects, Mild Cognitive Impairment (MCI), and Alzheimer's Disease (AD).

More and more large-scale imaging genetic studies are being widely conducted to collect a rich set of imaging, genetic, and clinical data to detect putative genes for complexly inherited neuropsychiatric and neurodegenerative disorders. Several major big-data challenges arise from testing genome-wide (NC>12 million known variants) associations with signals at millions of locations (NV~10(6)) in the brain from thousands of subjects (n~10(3)). The aim of this paper is to develop a Fast Voxelwise Genome Wide Association analysiS (FVGWAS) framework to efficiently carry out whole-genome analyses of whole-brain data. FVGWAS consists of three components including a heteroscedastic linear model, a global sure independence screening (GSIS) procedure, and a detection procedure based on wild bootstrap methods. Specifically, for standard linear association, the computational complexity is O (nNVNC) for voxelwise genome wide association analysis (VGWAS) method compared with O ((NC+NV)n(2)) for FVGWAS. Simulation studies show that FVGWAS is an efficient method of searching sparse signals in an extremely large search space, while controlling for the family-wise error rate. Finally, we have successfully applied FVGWAS to a large-scale imaging genetic data analysis of ADNI data with 708 subjects, 193,275voxels in RAVENS maps, and 501,584 SNPs, and the total processing time was 203,645s for a single CPU. Our FVGWAS may be a valuable statistical toolbox for large-scale imaging genetic analysis as the field is rapidly advancing with ultra-high-resolution imaging and whole-genome sequencing.

We investigated fluctuations in brain volume throughout the day using statistical modeling of magnetic resonance imaging (MRI) from large populations. We applied fully automated image analysis software to measure the brain parenchymal fraction (BPF), defined as the ratio of the brain parenchymal volume and intracranial volume, thus accounting for variations in head size. The MRI data came from serial scans of multiple sclerosis (MS) patients in clinical trials (n=755, 3269 scans) and from subjects participating in the Alzheimer's Disease Neuroimaging Initiative (ADNI, n=834, 6114 scans). The percent change in BPF was modeled with a linear mixed effect (LME) model, and the model was applied separately to the MS and ADNI datasets. The LME model for the MS datasets included random subject effects (intercept and slope over time) and fixed effects for the time-of-day, time from the baseline scan, and trial, which accounted for trial-related effects (for example, different inclusion criteria and imaging protocol). The model for ADNI additionally included the demographics (baseline age, sex, subject type [normal, mild cognitive impairment, or Alzheimer's disease], and interaction between subject type and time from baseline). There was a statistically significant effect of time-of-day on the BPF change in MS clinical trial datasets (-0.180 per day, that is, 0.180% of intracranial volume, p=0.019) as well as the ADNI dataset (-0.438 per day, that is, 0.438% of intracranial volume, p<0.0001), showing that the brain volume is greater in the morning. Linearly correcting the BPF values with the time-of-day reduced the required sample size to detect a 25% treatment effect (80% power and 0.05 significance level) on change in brain volume from 2 time-points over a period of 1year by 2.6%. Our results have significant implications for future brain volumetric studies, suggesting that there is a potential acquisition time bias that should be randomized or statistically controlled to account for the day-to-day brain volume fluctuations.

In this paper we present a method to segment four brainstem structures (midbrain, pons, medulla oblongata and superior cerebellar peduncle) from 3D brain MRI scans. The segmentation method relies on a probabilistic atlas of the brainstem and its neighboring brain structures. To build the atlas, we combined a dataset of 39 scans with already existing manual delineations of the whole brainstem and a dataset of 10 scans in which the brainstem structures were manually labeled with a protocol that was specifically designed for this study. The resulting atlas can be used in a Bayesian framework to segment the brainstem structures in novel scans. Thanks to the generative nature of the scheme, the segmentation method is robust to changes in MRI contrast or acquisition hardware. Using cross validation, we show that the algorithm can segment the structures in previously unseen T1 and FLAIR scans with great accuracy (mean error under 1mm) and robustness (no failures in 383 scans including 168 AD cases). We also indirectly evaluate the algorithm with a experiment in which we study the atrophy of the brainstem in aging. The results show that, when used simultaneously, the volumes of the midbrain, pons and medulla are significantly more predictive of age than the volume of the entire brainstem, estimated as their sum. The results also demonstrate that the method can detect atrophy patterns in the brainstem structures that have been previously described in the literature. Finally, we demonstrate that the proposed algorithm is able to detect differential effects of AD on the brainstem structures. The method will be implemented as part of the popular neuroimaging package FreeSurfer.

Tract-Based Spatial Statistics (TBSS) is a popular software pipeline to coregister sets of diffusion tensor Fractional Anisotropy (FA) images for performing voxel-wise comparisons. It is primarily defined by its skeleton projection step intended to reduce effects of local misregistration. A white matter "skeleton" is computed by morphological thinning of the inter-subject mean FA, and then all voxels are projected to the nearest location on this skeleton. Here we investigate several enhancements to the TBSS pipeline based on recent advances in registration for other modalities, principally based on groupwise registration with the ANTS-SyN algorithm. We validate these enhancements using simulation experiments with synthetically-modified images. When used with these enhancements, we discover that TBSS's skeleton projection step actually reduces algorithm accuracy, as the improved registration leaves fewer errors to warrant correction, and the effects of this projection's compromises become stronger than those of its benefits. In our experiments, our proposed pipeline without skeleton projection is more sensitive for detecting true changes and has greater specificity in resisting false positives from misregistration. We also present comparative results of the proposed and traditional methods, both with and without the skeleton projection step, on three real-life datasets: two comparing differing populations of Alzheimer's disease patients to matched controls, and one comparing progressive supranuclear palsy patients to matched controls. The proposed pipeline produces more plausible results according to each disease's pathophysiology.

Mild cognitive impairment (MCI) is a transitional stage between age-related cognitive decline and Alzheimer's disease (AD). For the effective treatment of AD, it would be important to identify MCI patients at high risk for conversion to AD. In this study, we present a novel magnetic resonance imaging (MRI)-based method for predicting the MCI-to-AD conversion from one to three years before the clinical diagnosis. First, we developed a novel MRI biomarker of MCI-to-AD conversion using semi-supervised learning and then integrated it with age and cognitive measures about the subjects using a supervised learning algorithm resulting in what we call the aggregate biomarker. The novel characteristics of the methods for learning the biomarkers are as follows: 1) We used a semi-supervised learning method (low density separation) for the construction of MRI biomarker as opposed to more typical supervised methods; 2) We performed a feature selection on MRI data from AD subjects and normal controls without using data from MCI subjects via regularized logistic regression; 3) We removed the aging effects from the MRI data before the classifier training to prevent possible confounding between AD and age related atrophies; and 4) We constructed the aggregate biomarker by first learning a separate MRI biomarker and then combining it with age and cognitive measures about the MCI subjects at the baseline by applying a random forest classifier. We experimentally demonstrated the added value of these novel characteristics in predicting the MCI-to-AD conversion on data obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. With the ADNI data, the MRI biomarker achieved a 10-fold cross-validated area under the receiver operating characteristic curve (AUC) of 0.7661 in discriminating progressive MCI patients (pMCI) from stable MCI patients (sMCI). Our aggregate biomarker based on MRI data together with baseline cognitive measurements and age achieved a 10-fold cross-validated AUC score of 0.9020 in discriminating pMCI from sMCI. The results presented in this study demonstrate the potential of the suggested approach for early AD diagnosis and an important role of MRI in the MCI-to-AD conversion prediction. However, it is evident based on our results that combining MRI data with cognitive test results improved the accuracy of the MCI-to-AD conversion prediction.

Proteomic and imaging markers have been widely studied as potential biomarkers for diagnosis, monitoring and prognosis of Alzheimer's disease. In this study, we used Alzheimer Disease Neuroimaging Initiative dataset and performed parallel independent component analysis on cross sectional and longitudinal proteomic and imaging data in order to identify the best proteomic model for diagnosis, monitoring and prediction of Alzheimer disease (AD). We used plasma proteins measurement and imaging data from AD and healthy controls (HC) at the baseline and 1 year follow-up. Group comparisons at baseline and changes over 1 year were calculated for proteomic and imaging data. The results were fed into parallel independent component analysis in order to identify proteins that were associated with structural brain changes cross sectionally and longitudinally. Regression model was used to find the best model that can discriminate AD from HC, monitor AD and to predict MCI converters from non-converters. We showed that five proteins are associated with structural brain changes in the brain. These proteins could discriminate AD from HC with 57% specificity and 89% sensitivity. Four proteins whose change over 1 year were associated with brain structural changes could discriminate AD from HC with sensitivity of 93%, and specificity of 92%. This model predicted MCI conversion to AD in 2 years with 94% accuracy. This model has the highest accuracy in prediction of MCI conversion to AD within the ADNI-1 dataset. This study shows that combination of selected plasma protein levels and MR imaging is a useful method in identifying potential biomarker.

Magnetic resonance imaging (MRI) intensities are acquired in arbitrary units, making scans non-comparable across sites and between subjects. Intensity normalization is a first step for the improvement of comparability of the images across subjects. However, we show that unwanted inter-scan variability associated with imaging site, scanner effect, and other technical artifacts is still present after standard intensity normalization in large multi-site neuroimaging studies. We propose RAVEL (Removal of Artificial Voxel Effect by Linear regression), a tool to remove residual technical variability after intensity normalization. As proposed by SVA and RUV [Leek and Storey, 2007, 2008, Gagnon-Bartsch and Speed, 2012], two batch effect correction tools largely used in genomics, we decompose the voxel intensities of images registered to a template into a biological component and an unwanted variation component. The unwanted variation component is estimated from a control region obtained from the cerebrospinal fluid (CSF), where intensities are known to be unassociated with disease status and other clinical covariates. We perform a singular value decomposition (SVD) of the control voxels to estimate factors of unwanted variation. We then estimate the unwanted factors using linear regression for every voxel of the brain and take the residuals as the RAVEL-corrected intensities. We assess the performance of RAVEL using T1-weighted (T1-w) images from more than 900 subjects with Alzheimer's disease (AD) and mild cognitive impairment (MCI), as well as healthy controls from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. We compare RAVEL to two intensity-normalization-only methods: histogram matching and White Stripe. We show that RAVEL performs best at improving the replicability of the brain regions that are empirically found to be most associated with AD, and that these regions are significantly more present in structures impacted by AD (hippocampus, amygdala, parahippocampal gyrus, enthorinal area, and fornix stria terminals). In addition, we show that the RAVEL-corrected intensities have the best performance in distinguishing between MCI subjects and healthy subjects using the mean hippocampal intensity (AUC=67%), a marked improvement compared to results from intensity normalization alone (AUC=63% and 59% for histogram matching and White Stripe, respectively). RAVEL is promising for many other imaging modalities.

The human hippocampal formation can be divided into a set of cytoarchitecturally and functionally distinct subregions, involved in different aspects of memory formation. Neuroanatomical disruptions within these subregions are associated with several debilitating brain disorders including Alzheimer's disease, major depression, schizophrenia, and bipolar disorder. Multi-center brain imaging consortia, such as the Enhancing Neuro Imaging Genetics through Meta-Analysis (ENIGMA) consortium, are interested in studying disease effects on these subregions, and in the genetic factors that affect them. For large-scale studies, automated extraction and subsequent genomic association studies of these hippocampal subregion measures may provide additional insight. Here, we evaluated the test-retest reliability and transplatform reliability (1.5T versus 3T) of the subregion segmentation module in the FreeSurfer software package using three independent cohorts of healthy adults, one young (Queensland Twins Imaging Study, N=39), another elderly (Alzheimer's Disease Neuroimaging Initiative, ADNI-2, N=163) and another mixed cohort of healthy and depressed participants (Max Planck Institute, MPIP, N=598). We also investigated agreement between the most recent version of this algorithm (v6.0) and an older version (v5.3), again using the ADNI-2 and MPIP cohorts in addition to a sample from the Netherlands Study for Depression and Anxiety (NESDA) (N=221). Finally, we estimated the heritability (h(2)) of the segmented subregion volumes using the full sample of young, healthy QTIM twins (N=728). Test-retest reliability was high for all twelve subregions in the 3T ADNI-2 sample (intraclass correlation coefficient (ICC)=0.70-0.97) and moderate-to-high in the 4T QTIM sample (ICC=0.5-0.89). Transplatform reliability was strong for eleven of the twelve subregions (ICC=0.66-0.96); however, the hippocampal fissure was not consistently reconstructed across 1.5T and 3T field strengths (ICC=0.47-0.57). Between-version agreement was moderate for the hippocampal tail, subiculum and presubiculum (ICC=0.78-0.84; Dice Similarity Coefficient (DSC)=0.55-0.70), and poor for all other subregions (ICC=0.34-0.81; DSC=0.28-0.51). All hippocampal subregion volumes were highly heritable (h(2)=0.67-0.91). Our findings indicate that eleven of the twelve human hippocampal subregions segmented using FreeSurfer version 6.0 may serve as reliable and informative quantitative phenotypes for future multi-site imaging genetics initiatives such as those of the ENIGMA consortium.

Brain extraction is an important step in the analysis of brain images. The variability in brain morphology and the difference in intensity characteristics due to imaging sequences make the development of a general purpose brain extraction algorithm challenging. To address this issue, we propose a new robust method (BEaST) dedicated to produce consistent and accurate brain extraction. This method is based on nonlocal segmentation embedded in a multi-resolution framework. A library of 80 priors is semi-automatically constructed from the NIH-sponsored MRI study of normal brain development, the International Consortium for Brain Mapping, and the Alzheimer's Disease Neuroimaging Initiative databases. In testing, a mean Dice similarity coefficient of 0.9834±0.0053 was obtained when performing leave-one-out cross validation selecting only 20 priors from the library. Validation using the online Segmentation Validation Engine resulted in a top ranking position with a mean Dice coefficient of 0.9781±0.0047. Robustness of BEaST is demonstrated on all baseline ADNI data, resulting in a very low failure rate. The segmentation accuracy of the method is better than two widely used publicly available methods and recent state-of-the-art hybrid approaches. BEaST provides results comparable to a recent label fusion approach, while being 40 times faster and requiring a much smaller library of priors.

Longitudinal MRI studies are often subjected to mid-study scanner changes, which may alter image characteristics such as contrast, signal-to-noise ratio, contrast-to-noise ratio, intensity non-uniformity and geometric distortion. Measuring brain volume loss under these conditions can render the results potentially unreliable across the timepoint of the change. Estimating and accounting for this effect may improve the reliability of estimates of brain atrophy rates.

In this study we propose a deformation-based framework to jointly model the influence of aging and Alzheimer's disease (AD) on the brain morphological evolution. Our approach combines a spatio-temporal description of both processes into a generative model. A reference morphology is deformed along specific trajectories to match subject specific morphologies. It is used to define two imaging progression markers: 1) a morphological age and 2) a disease score. These markers can be computed regionally in any brain region. The approach is evaluated on brain structural magnetic resonance images (MRI) from the ADNI database. The model is first estimated on a control population using longitudinal data, then, for each testing subject, the markers are computed cross-sectionally for each acquisition. The longitudinal evolution of these markers is then studied in relation with the clinical diagnosis of the subjects and used to generate possible morphological evolutions. In the model, the morphological changes associated with normal aging are mainly found around the ventricles, while the Alzheimer's disease specific changes are located in the temporal lobe and the hippocampal area. The statistical analysis of these markers highlights differences between clinical conditions even though the inter-subject variability is quite high. The model is also generative since it can be used to simulate plausible morphological trajectories associated with the disease. Our method quantifies two interpretable scalar imaging biomarkers assessing respectively the effects of aging and disease on brain morphology, at the individual and population level. These markers confirm the presence of an accelerated apparent aging component in Alzheimer's patients but they also highlight specific morphological changes that can help discriminate clinical conditions even in prodromal stages. More generally, the joint modeling of normal and pathological evolutions shows promising results to describe age-related brain diseases over long time scales.

Alzheimer's disease (AD) is a debilitating age-related neurodegenerative disorder. Accurate identification of individuals at risk is complicated as AD shares cognitive and brain features with aging. We applied linked independent component analysis (LICA) on three complementary measures of gray matter structure: cortical thickness, area and gray matter density of 137 AD, 78 mild (MCI) and 38 subjective cognitive impairment patients, and 355 healthy adults aged 18-78 years to identify dissociable multivariate morphological patterns sensitive to age and diagnosis. Using the lasso classifier, we performed group classification and prediction of cognition and age at different age ranges to assess the sensitivity and diagnostic accuracy of the LICA patterns in relation to AD, as well as early and late healthy aging. Three components showed high sensitivity to the diagnosis and cognitive status of AD, with different relationships with age: one reflected an anterior-posterior gradient in thickness and gray matter density and was uniquely related to diagnosis, whereas the other two, reflecting widespread cortical thickness and medial temporal lobe volume, respectively, also correlated significantly with age. Repeating the LICA decomposition and between-subject analysis on ADNI data, including 186 AD, 395 MCI and 220 age-matched healthy controls, revealed largely consistent brain patterns and clinical associations across samples. Classification results showed that multivariate LICA-derived brain characteristics could be used to predict AD and age with high accuracy (area under ROC curve up to 0.93 for classification of AD from controls). Comparison between classifiers based on feature ranking and feature selection suggests both common and unique feature sets implicated in AD and aging, and provides evidence of distinct age-related differences in early compared to late aging.

Early identification of individuals at risk of developing Alzheimer's disease (AD) dementia is important for developing disease-modifying therapies. In this study, given multimodal AD markers and clinical diagnosis of an individual from one or more timepoints, we seek to predict the clinical diagnosis, cognition and ventricular volume of the individual for every month (indefinitely) into the future. We proposed and applied a minimal recurrent neural network (minimalRNN) model to data from The Alzheimer's Disease Prediction Of Longitudinal Evolution (TADPOLE) challenge, comprising longitudinal data of 1677 participants (Marinescu et al., 2018) from the Alzheimer's Disease Neuroimaging Initiative (ADNI). We compared the performance of the minimalRNN model and four baseline algorithms up to 6 years into the future. Most previous work on predicting AD progression ignore the issue of missing data, which is a prevalent issue in longitudinal data. Here, we explored three different strategies to handle missing data. Two of the strategies treated the missing data as a "preprocessing" issue, by imputing the missing data using the previous timepoint ("forward filling") or linear interpolation ("linear filling). The third strategy utilized the minimalRNN model itself to fill in the missing data both during training and testing ("model filling"). Our analyses suggest that the minimalRNN with "model filling" compared favorably with baseline algorithms, including support vector machine/regression, linear state space (LSS) model, and long short-term memory (LSTM) model. Importantly, although the training procedure utilized longitudinal data, we found that the trained minimalRNN model exhibited similar performance, when using only 1 input timepoint or 4 input timepoints, suggesting that our approach might work well with just cross-sectional data. An earlier version of our approach was ranked 5th (out of 53 entries) in the TADPOLE challenge in 2019. The current approach is ranked 2nd out of 63 entries as of June 3rd, 2020.

Current theory suggests brain regions interact to reconcile the competing demands of integration and segregation by leveraging metastable dynamics. An emerging consensus recognises the importance of metastability in healthy neural dynamics where the transition between network states over time is dependent upon the structural connectivity between brain regions. In Alzheimer's disease (AD) - the most common form of dementia - these couplings are progressively weakened, metastability of neural dynamics are reduced and cognitive ability is impaired. Accordingly, we use a joint empirical and computational approach to reveal how behaviourally relevant changes in neural metastability are contingent on the structural integrity of the anatomical connectome. We estimate the metastability of fMRI BOLD signal in subjects from across the AD spectrum and in healthy controls and demonstrate the dissociable effects of structural disconnection on synchrony versus metastability. In addition, we reveal the critical role of metastability in general cognition by demonstrating the link between an individuals cognitive performance and their metastable neural dynamic. Finally, using whole-brain computer modelling, we demonstrate how a healthy neural dynamic is conditioned upon the topological integrity of the structural connectome. Overall, the results of our joint computational and empirical analysis suggest an important causal relationship between metastable neural dynamics, cognition, and the structural efficiency of the anatomical connectome.