AddNeuroMed and ADNI: similar patterns of Alzheimer's atrophy and automated MRI classification accuracy in Europe and North America

Predicting progression from subjective cognitive decline to mild cognitive impairment or dementia based on brain atrophy patterns

BACKGROUND

Alzheimer's disease (AD) is a progressive neurodegenerative disorder where pathophysiological changes begin decades before the onset of clinical symptoms. Analysis of brain atrophy patterns using structural MRI and multivariate data analysis are an effective tool in identifying patients with subjective cognitive decline (SCD) at higher risk of progression to AD dementia. Atrophy patterns obtained from models trained to classify advanced AD versus normal subjects, may not be optimal for subjects at an early stage, like SCD. In this study, we compared the accuracy of the SCD progression prediction using the 'severity index' generated using a standard classification model trained on patients with AD dementia versus a new model trained on β-amyloid (Aβ) positive patients with amnestic mild cognitive impairment (aMCI).

METHODS

We used structural MRI data of 504 patients from the Swedish BioFINDER-1 study cohort (cognitively normal (CN), Aβ-negative = 220; SCD, Aβ positive and negative = 139; aMCI, Aβ-positive = 106; AD dementia = 39). We applied multivariate data analysis to create two predictive models trained to discriminate CN individuals from either individuals with Aβ positive aMCI or AD dementia. Models were applied to individuals with SCD to classify their atrophy patterns as either high-risk "disease-like" or low-risk "CN-like". Clinical trajectory and model accuracy were evaluated using 8 years of longitudinal data.

RESULTS

In predicting progression from SCD to MCI or dementia, the standard, dementia-based model, reached 100% specificity but only 10.6% sensitivity, while the new, aMCI-based model, reached 72.3% sensitivity and 60.9% specificity. The aMCI-based model was superior in predicting progression from SCD to MCI or dementia, reaching a higher receiver operating characteristic area under curve (AUC = 0.72; P = 0.037) in comparison with the dementia-based model (AUC = 0.57).

CONCLUSION

When predicting conversion from SCD to MCI or dementia using structural MRI data, prediction models based on individuals with milder levels of atrophy (i.e. aMCI) may offer superior clinical value compared to standard dementia-based models.

Hippocampal transcriptome-wide association study and pathway analysis of mitochondrial solute carriers in Alzheimer's disease

The etiopathogenesis of late-onset Alzheimer's disease (AD) is increasingly recognized as the result of the combination of the aging process, toxic proteins, brain dysmetabolism, and genetic risks. Although the role of mitochondrial dysfunction in the pathogenesis of AD has been well-appreciated, the interaction between mitochondrial function and genetic variability in promoting dementia is still poorly understood. In this study, by tissue-specific transcriptome-wide association study (TWAS) and further meta-analysis, we examined the genetic association between mitochondrial solute carrier family (SLC25) genes and AD in three independent cohorts and identified three AD-susceptibility genes, including SLC25A10, SLC25A17, and SLC25A22. Integrative analysis using neuroimaging data and hippocampal TWAS-predicted gene expression of the three susceptibility genes showed an inverse correlation of SLC25A22 with hippocampal atrophy rate in AD patients, which outweighed the impacts of sex, age, and apolipoprotein E4 (ApoE4). Furthermore, SLC25A22 downregulation demonstrated an association with AD onset, as compared with the other two transcriptome-wide significant genes. Pathway and network analysis related hippocampal SLC25A22 downregulation to defects in neuronal function and development, echoing the enrichment of SLC25A22 expression in human glutamatergic neurons. The most parsimonious interpretation of the results is that we have identified AD-susceptibility genes in the SLC25 family through the prediction of hippocampal gene expression. Moreover, our findings mechanistically yield insight into the mitochondrial cascade hypothesis of AD and pave the way for the future development of diagnostic tools for the early prevention of AD from a perspective of precision medicine by targeting the mitochondria-related genes.

Protective Effects of Whey Protein Hydrolysate, Treadmill Exercise, and Their Combination against Scopolamine-Induced Cognitive Deficit in Mice

In this study, the potential of whey protein hydrolysate (WPH) and treadmill exercise to prevent cognitive decline was investigated, along with their neuroprotective mechanisms. Cognitive dysfunction was induced in mice with 1 mg/kg of scopolamine, followed by the administration of WPH at 100 and 200 mg/kg and/or treadmill exercise at 15 m/min for 30 min five days per week. Both WPH administration and treadmill exercise significantly improved the memory of mice with scopolamine-induced cognitive impairment, which was attributed to several key mechanisms, including a reduction in oxidative stress based on decreased levels of reactive oxygen species and malondialdehyde in the brain tissue and an increase in acetylcholine by increasing choline acyltransferase and decreasing acetylcholine esterase levels. Exercise and WPH also exerted neuroprotective effects by inhibiting the hyperphosphorylation of tau proteins, enhancing the expression of the brain-derived neurotrophic factor, and inhibiting apoptosis by reducing the Bax/Bcl2 ratio in conjunction with the downregulation of the mitogen-activated protein kinase pathway. Moreover, the impact of WPH and treadmill exercise extended to the gut microbiome, suggesting a potential link with cognitive improvement. These findings suggest that both WPH intake and treadmill exercise are effective strategies for mitigating cognitive impairment, providing promising avenues for treating neurodegenerative diseases.

A robust class decomposition-based approach for detecting Alzheimer's progression

Computer-aided diagnosis of Alzheimer's disease (AD) is a rapidly growing field with the possibility to be utilized in practice. Deep learning has received much attention in detecting AD from structural magnetic resonance imaging (sMRI). However, training a convolutional neural network from scratch is problematic because it requires a lot of annotated data and additional computational time. Transfer learning can offer a promising and practical solution by transferring information learned from other image recognition tasks to medical image classification. Another issue is the dataset distribution's irregularities. A common classification issue in datasets is a class imbalance, where the distribution of samples among the classes is biased. For example, a dataset may contain more instances of some classes than others. Class imbalance is challenging because most machine learning algorithms assume that each class should have an equal number of samples. Models consequently perform poorly in prediction. Class decomposition can address this problem by making learning a dataset's class boundaries easier. Motivated by these approaches, we propose a class decomposition transfer learning (CDTL) approach that employs VGG19, AlexNet, and an entropy-based technique to detect AD from sMRI. This study aims to assess the robustness of the CDTL approach in detecting the cognitive decline of AD using data from various ADNI cohorts to determine whether comparable classification accuracy for the two or more cohorts would be obtained. Furthermore, the proposed model achieved state-of-the-art performance in predicting mild cognitive impairment (MCI)-to-AD conversion with an accuracy of 91.45%.

Distinct sex-specific DNA methylation differences in Alzheimer's disease

BACKGROUND

Sex is increasingly recognized as a significant factor contributing to the biological and clinical heterogeneity in AD. There is also growing evidence for the prominent role of DNA methylation (DNAm) in Alzheimer's disease (AD).

METHODS

We studied sex-specific DNA methylation differences in the blood samples of AD subjects compared to cognitively normal subjects, by performing sex-specific meta-analyses of two large blood-based epigenome-wide association studies (ADNI and AIBL), which included DNA methylation data for a total of 1284 whole blood samples (632 females and 652 males). Within each dataset, we used two complementary analytical strategies, a sex-stratified analysis that examined methylation to AD associations in male and female samples separately, and a methylation-by-sex interaction analysis that compared the magnitude of these associations between different sexes. After adjusting for age, estimated immune cell type proportions, batch effects, and correcting for inflation, the inverse-variance fixed-effects meta-analysis model was used to identify the most consistent DNAm differences across datasets. In addition, we also evaluated the performance of the sex-specific methylation-based risk prediction models for AD diagnosis using an independent external dataset.

RESULTS

In the sex-stratified analysis, we identified 2 CpGs, mapped to the PRRC2A and RPS8 genes, significantly associated with AD in females at a 5% false discovery rate, and an additional 25 significant CpGs (21 in females, 4 in males) at P-value < 1×10^-5. In methylation-by-sex interaction analysis, we identified 5 significant CpGs at P-value < 10^-5. Out-of-sample validations using the AddNeuroMed dataset showed in females, the best logistic prediction model included age, estimated immune cell-type proportions, and methylation risk scores (MRS) computed from 9 of the 23 CpGs identified in AD vs. CN analysis that are also available in AddNeuroMed dataset (AUC = 0.74, 95% CI: 0.65-0.83). In males, the best logistic prediction model included only age and MRS computed from 2 of the 5 CpGs identified in methylation-by-sex interaction analysis that are also available in the AddNeuroMed dataset (AUC = 0.70, 95% CI: 0.56-0.82).

CONCLUSIONS

Overall, our results show that the DNA methylation differences in AD are largely distinct between males and females. Our best-performing sex-specific methylation-based prediction model in females performed better than that for males and additionally included estimated cell-type proportions. The significant discriminatory classification of AD samples with our methylation-based prediction models demonstrates that sex-specific DNA methylation could be a predictive biomarker for AD. As sex is a strong factor underlying phenotypic variability in AD, the results of our study are particularly relevant for a better understanding of the epigenetic architecture that underlie AD and for promoting precision medicine in AD.

Classification of Mild Cognitive Impairment With Multimodal Data Using Both Labeled and Unlabeled Samples

Mild Cognitive Impairment (MCI) is a preclinical stage of Alzheimer's Disease (AD) and is clinical heterogeneity. The classification of MCI is crucial for the early diagnosis and treatment of AD. In this study, we investigated the potential of using both labeled and unlabeled samples from the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort to classify MCI through the multimodal co-training method. We utilized both structural magnetic resonance imaging (sMRI) data and genotype data of 364 MCI samples including 228 labeled and 136 unlabeled MCI samples from the ADNI-1 cohort. First, the selected quantitative trait (QT) features from sMRI data and SNP features from genotype data were used to build two initial classifiers on 228 labeled MCI samples. Then, the co-training method was implemented to obtain new labeled samples from 136 unlabeled MCI samples. Finally, the random forest algorithm was used to obtain a combined classifier to classify MCI patients in the independent ADNI-2 dataset. The experimental results showed that our proposed framework obtains an accuracy of 85.50 percent and an AUC of 0.825 for MCI classification, respectively, which showed that the combined utilization of sMRI and SNP data through the co-training method could significantly improve the performances of MCI classification.

Long-term follow-up of neurodegenerative phenomenon in severe traumatic brain injury using MRI

BACKGROUND

Traumatic brain injury (TBI) lesions are known to evolve over time, but the duration and consequences of cerebral remodeling are unclear. Degenerative mechanisms occurring in the chronic phase after TBI could constitute "tertiary" lesions related to the neurological outcome.

OBJECTIVE

The objective of this prospective study of severe TBI was to longitudinally evaluate the volume of white and grey matter structures and white matter integrity with 2 time-point multimodal MRI.

METHODS

Longitudinal MRI follow-up was obtained for 11 healthy controls (HCs) and 22 individuals with TBI (mean [SD] 60 [15] months after injury) along with neuropsychological assessments. TBI individuals were classified in the "favourable" recovery group (Glasgow Outcome Scale Extended [GOSE] 6-8) and "unfavourable" recovery group (GOSE 3-5) at 5 years. Variation in brain volumes (3D T1-weighted image) and white matter integrity (diffusion tensor imaging [DTI]) were quantitatively assessed over time and used to predict neurological outcome.

RESULTS

TBI individuals showed a marked decrease in volumes of whole white matter (median -11.4% [interquartile range -5.8; -14.6]; p <0.001) and deep grey nuclear structures (-17.1% [-10.6; -20.5]; p <0.001). HCs did not show any significant change over the same time period. Median volumetric loss in several brain regions was higher with GOSE 3-5 than 6-8. These lesions were associated with lower fractional anisotropy and higher mean diffusivity at baseline. Volumetric variations were positively correlated with normalized fractional anisotropy and negatively with normalized mean diffusivity at baseline and follow-up. A computed predictive model with baseline DTI showed good accuracy to predict neurological outcome (area under the receiver operating characteristic curve 0.82 [95% confidence interval 0.81-0.83]) Conclusions. We characterised the striking atrophy of deep brain structures after severe TBI. DTI imaging in the subacute phase can predict the occurrence and localization of these tertiary lesions as well as long-term neurological outcome.

TRIAL REGISTRATION

ClinicalTrials.gov: NCT00577954. Registered on October 2006.

Influence of White Matter Hyperintensities on Baseline and Longitudinal Amyloid-β in Cognitively Normal Individuals

BACKGROUND

The associations between small vessel disease (SVD) and cerebrospinal amyloid-β1-42 (Aβ1-42) pathology have not been well-elucidated.

OBJECTIVE

Baseline (BL) white matter hyperintensities (WMH) were examined for associations with month-24 (M24) and longitudinal Aβ1-42 change in cognitively normal (CN) subjects. The interaction of WMH and Aβ1-42 on memory and executive function were also examined.

METHODS

This study included 72 subjects from the Alzheimer's Disease Neuroimaging Initiative. Multivariable linear regression models evaluated associations between baseline WMH/intracranial volume ratio, M24 and change in Aβ1-42 over two years. Linear mixed effects models evaluated interactions between BL WMH/ICV and Aβ1-42 on memory and executive function.

RESULTS

Mean age of the subjects (Nmales = 36) = 73.80 years, SD = 6.73; mean education years = 17.1, SD = 2.4. BL WMH was significantly associated with M24 Aβ1-42 (p = 0.008) and two-year change in Aβ1-42 (p = 0.006). Interaction between higher WMH and lower Aβ1-42 at baseline was significantly associated with worse memory at baseline and M24 (p = 0.003).

CONCLUSION

BL WMH was associated with M24 and longitudinal Aβ1-42 change in CN. The interaction between higher WMH and lower Aβ1-42 was associated with poorer memory. Since SVD is associated with longitudinal Aβ1-42 pathology, and the interaction of both factors is linked to poorer cognitive outcomes, the mitigation of SVD may be correlated with reduced amyloid pathology and milder cognitive deterioration in Alzheimer's disease.

An epigenome-wide association study of Alzheimer's disease blood highlights robust DNA hypermethylation in the HOXB6 gene

A growing number of epigenome-wide association studies have demonstrated a role for DNA methylation in the brain in Alzheimer's disease. With the aim of exploring peripheral biomarker potential, we have examined DNA methylation patterns in whole blood collected from 284 individuals in the AddNeuroMed study, which included 89 nondemented controls, 86 patients with Alzheimer's disease, and 109 individuals with mild cognitive impairment, including 38 individuals who progressed to Alzheimer's disease within 1 year. We identified significant differentially methylated regions, including 12 adjacent hypermethylated probes in the HOXB6 gene in Alzheimer's disease, which we validated using pyrosequencing. Using weighted gene correlation network analysis, we identified comethylated modules of genes that were associated with key variables such as APOE genotype and diagnosis. In summary, this study represents the first large-scale epigenome-wide association study of Alzheimer's disease and mild cognitive impairment using blood. We highlight the differences in various loci and pathways in early disease, suggesting that these patterns relate to cognitive decline at an early stage.

Genome-wide transcriptome analysis identifies novel dysregulated genes implicated in Alzheimer's pathology

INTRODUCTION

Abnormal gene expression patterns may contribute to the onset and progression of late-onset Alzheimer's disease (LOAD).

METHODS

We performed transcriptome-wide meta-analysis (N = 1440) of blood-based microarray gene expression profiles as well as neuroimaging and cerebrospinal fluid (CSF) endophenotype analysis.

RESULTS

We identified and replicated five genes (CREB5, CD46, TMBIM6, IRAK3, and RPAIN) as significantly dysregulated in LOAD. The most significantly altered gene, CREB5, was also associated with brain atrophy and increased amyloid beta (Aβ) accumulation, especially in the entorhinal cortex region. cis-expression quantitative trait loci mapping analysis of CREB5 detected five significant associations (P < 5 × 10^-8 ), where rs56388170 (most significant) was also significantly associated with global cortical Aβ deposition measured by [¹⁸ F]Florbetapir positron emission tomography and CSF Aβ_1-42 .

DISCUSSION

RNA from peripheral blood indicated a differential gene expression pattern in LOAD. Genes identified have been implicated in biological processes relevant to Alzheimer's disease. CREB, in particular, plays a key role in nervous system development, cell survival, plasticity, and learning and memory.

Comparative Utility of Manual versus Automated Segmentation of Hippocampus and Entorhinal Cortex Volumes in a Memory Clinic Sample

Structural neuroimaging is a useful non-invasive biomarker commonly employed to evaluate the integrity of mesial temporal lobe structures that are typically compromised in Alzheimer's disease. Advances in quantitative neuroimaging have permitted the development of automated segmentation protocols (e.g., FreeSurfer) with significantly increased efficiency compared to earlier manual techniques. While these protocols have been found to be suitable for large-scale, multi-site research studies, we were interested in assessing the practical utility and reliability of automated FreeSurfer protocols compared to manual volumetry on routinely acquired clinical scans. Independent validation studies with newer automated segmentation protocols are scarce. Two FreeSurfer protocols for each of two regions of interest-the hippocampus and entorhinal cortex-were compared against manual volumetry. High reliability and agreement was found between FreeSurfer and manual hippocampal protocols, however, there was lower reliability and agreement between FreeSurfer and manual entorhinal protocols. Although based on a the relatively small sample of subjects drawn from a memory clinic (n = 27), our study findings suggest further refinements to improve measurement error and most accurately depict true regional brain volumes using automated segmentation protocols are required, especially for non-hippocampal mesial temporal structures, to achieve maximal utility for routine clinical evaluations.

Predicting progression from mild cognitive impairment to Alzheimer's disease on an individual subject basis by applying the CARE index across different independent cohorts

The purposes of this study are to investigate whether the Characterizing Alzheimer's disease Risk Events (CARE) index can accurately predict progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD) on an individual subject basis, and to investigate whether this model can be generalized to an independent cohort. Using an event-based probabilistic model approach to integrate widely available biomarkers from behavioral data and brain structural and functional imaging, we calculated the CARE index. We then applied the CARE index to identify which MCI individuals from the ADNI dataset progressed to AD during a three-year follow-up period. Subsequently, the CARE index was generalized to the prediction of MCI individuals from an independent Nanjing Aging and Dementia Study (NADS) dataset during the same time period. The CARE index achieved high prediction performance with 80.4% accuracy, 75% sensitivity, 82% specificity, and 0.809 area under the receiver operating characteristic (ROC) curve (AUC) on MCI subjects from the ADNI dataset over three years, and a highly validated prediction performance with 87.5% accuracy, 81% sensitivity, 90% specificity, and 0.861 AUC on MCI subjects from the NADS dataset. In conclusion, the CARE index is highly accurate, sufficiently robust, and generalized for predicting which MCI individuals will develop AD over a three-year period. This suggests that the CARE index can be usefully applied to select individuals with MCI for clinical trials and to identify which individuals will convert from MCI to AD for administration of early disease-modifying treatment.

The reliability of a deep learning model in clinical out-of-distribution MRI data: A multicohort study

Deep learning (DL) methods have in recent years yielded impressive results in medical imaging, with the potential to function as clinical aid to radiologists. However, DL models in medical imaging are often trained on public research cohorts with images acquired with a single scanner or with strict protocol harmonization, which is not representative of a clinical setting. The aim of this study was to investigate how well a DL model performs in unseen clinical datasets-collected with different scanners, protocols and disease populations-and whether more heterogeneous training data improves generalization. In total, 3117 MRI scans of brains from multiple dementia research cohorts and memory clinics, that had been visually rated by a neuroradiologist according to Scheltens' scale of medial temporal atrophy (MTA), were included in this study. By training multiple versions of a convolutional neural network on different subsets of this data to predict MTA ratings, we assessed the impact of including images from a wider distribution during training had on performance in external memory clinic data. Our results showed that our model generalized well to datasets acquired with similar protocols as the training data, but substantially worse in clinical cohorts with visibly different tissue contrasts in the images. This implies that future DL studies investigating performance in out-of-distribution (OOD) MRI data need to assess multiple external cohorts for reliable results. Further, by including data from a wider range of scanners and protocols the performance improved in OOD data, which suggests that more heterogeneous training data makes the model generalize better. To conclude, this is the most comprehensive study to date investigating the domain shift in deep learning on MRI data, and we advocate rigorous evaluation of DL models on clinical data prior to being certified for deployment.

Application of Generalized Split Linearized Bregman Iteration algorithm for Alzheimer's disease prediction

In this paper, we applied a novel method for the detection of Alzheimer's disease (AD) based on a structural magnetic resonance imaging (sMRI) dataset. Specifically, the method involved a new classification algorithm of machine learning, named Generalized Split Linearized Bregman Iteration (GSplit LBI). It combines logistic regression and structural sparsity regularizations. In the study, 57 AD patients and 47 normal controls (NCs) were enrolled. We first extracted the entire brain gray matter volume values of all subjects and then used GSplit LBI to build a predictive classification model with a 10-fold full cross-validation method. The model accuracy achieved 90.44%. To further verify which voxels in the dataset have greater impact on the prediction results, we ranked the weight parameters and obtained the top 6% of the model parameters. To verify the generalization of model prediction and the stability of feature selection, we performed a cross-test on the Alzheimer's Disease Neuroimaging Initiative (ADNI) and a Chinese dataset and achieved good performances on different cohorts. Conclusively, based on the sMRI dataset, our algorithm not only had good performance in a local cohort with high accuracy but also had good generalization of model prediction and stability of feature selection in different cohorts.

Machine Learning Techniques for the Diagnosis of Alzheimer’s Disease

Alzheimer’s disease is an incurable neurodegenerative disease primarily affecting the elderly population. Efficient automated techniques are needed for early diagnosis of Alzheimer’s. Many novel approaches are proposed by researchers for classification of Alzheimer’s disease. However, to develop more efficient learning techniques, better understanding of the work done on Alzheimer’s is needed. Here, we provide a review on 165 papers from 2005 to 2019, using various feature extraction and machine learning techniques. The machine learning techniques are surveyed under three main categories: support vector machine (SVM), artificial neural network (ANN), and deep learning (DL) and ensemble methods. We present a detailed review on these three approaches for Alzheimer’s with possible future directions.

Shape Information Improves the Cross-Cohort Performance of Deep Learning-Based Segmentation of the Hippocampus

Performing an accurate segmentation of the hippocampus from brain magnetic resonance images is a crucial task in neuroimaging research, since its structural integrity is strongly related to several neurodegenerative disorders, including Alzheimer's disease (AD). Some automatic segmentation tools are already being used, but, in recent years, new deep learning (DL)-based methods have been proven to be much more accurate in various medical image segmentation tasks. In this work, we propose a DL-based hippocampus segmentation framework that embeds statistical shape of the hippocampus as context information into the deep neural network (DNN). The inclusion of shape information is achieved with three main steps: (1) a U-Net-based segmentation, (2) a shape model estimation, and (3) a second U-Net-based segmentation which uses both the original input data and the fitted shape model. The trained DL architectures were tested on image data of three diagnostic groups [AD patients, subjects with mild cognitive impairment (MCI) and controls] from two cohorts (ADNI and AddNeuroMed). Both intra-cohort validation and cross-cohort validation were performed and compared with the conventional U-net architecture and some variations with other types of context information (i.e., autocontext and tissue-class context). Our results suggest that adding shape information can improve the segmentation accuracy in cross-cohort validation, i.e., when DNNs are trained on one cohort and applied to another. However, no significant benefit is observed in intra-cohort validation, i.e., training and testing DNNs on images from the same cohort. Moreover, compared to other types of context information, the use of shape context was shown to be the most successful in increasing the accuracy, while keeping the computational time in the order of a few minutes.

Dickkopf-1 Overexpression in vitro Nominates Candidate Blood Biomarkers Relating to Alzheimer's Disease Pathology

BACKGROUND

Previous studies suggest that Dickkopf-1 (DKK1), an inhibitor of Wnt signaling, plays a role in amyloid-induced toxicity and hence Alzheimer's disease (AD). However, the effect of DKK1 expression on protein expression, and whether such proteins are altered in disease, is unknown.

OBJECTIVE

We aim to test whether DKK1 induced protein signature obtained in vitro were associated with markers of AD pathology as used in the amyloid/tau/neurodegeneration (ATN) framework as well as with clinical outcomes.

METHODS

We first overexpressed DKK1 in HEK293A cells and quantified 1,128 proteins in cell lysates using aptamer capture arrays (SomaScan) to obtain a protein signature induced by DKK1. We then used the same assay to measure the DKK1-signature proteins in human plasma in two large cohorts, EMIF (n = 785) and ANM (n = 677).

RESULTS

We identified a 100-protein signature induced by DKK1 in vitro. Subsets of proteins, along with age and apolipoprotein E ɛ4 genotype distinguished amyloid pathology (A + T-N-, A+T+N-, A+T-N+, and A+T+N+) from no AD pathology (A-T-N-) with an area under the curve of 0.72, 0.81, 0.88, and 0.85, respectively. Furthermore, we found that some signature proteins (e.g., Complement C3 and albumin) were associated with cognitive score and AD diagnosis in both cohorts.

CONCLUSIONS

Our results add further evidence for a role of DKK regulation of Wnt signaling in AD and suggest that DKK1 induced signature proteins obtained in vitro could reflect theATNframework as well as predict disease severity and progression in vivo.

Using path signatures to predict a diagnosis of Alzheimer's disease

The path signature is a means of feature generation that can encode nonlinear interactions in data in addition to the usual linear terms. It provides interpretable features and its output is a fixed length vector irrespective of the number of input points or their sample times. In this paper we use the path signature to provide features for identifying people whose diagnosis subsequently converts to Alzheimer's disease. In two separate classification tasks we distinguish converters from 1) healthy individuals, and 2) individuals with mild cognitive impairment. The data used are time-ordered measurements of the whole brain, ventricles and hippocampus from the Alzheimer's Disease Neuroimaging Initiative (ADNI). We find two nonlinear interactions which are predictive in both cases. The first interaction is change of hippocampal volume with time, and the second is a change of hippocampal volume relative to the volume of the whole brain. While hippocampal and brain volume changes are well known in Alzheimer's disease, we demonstrate the power of the path signature in their identification and analysis without manual feature selection. Sequential data is becoming increasingly available as monitoring technology is applied, and the path signature method is shown to be a useful tool in the processing of this data.

Distinct structural correlates of the dominant and nondominant languages in bilinguals with Alzheimer's disease (AD)

Structural adaptations in brain regions involved in domain-general cognitive control are associated with life-long bilingualism and may contribute to the executive function advantage of bilinguals over monolinguals. To the degree that these adaptations support bilingualism, their disruption by Alzheimer's disease (AD) may compromise the ability to maintain proficiency in two languages, particularly in the less proficient, or nondominant, language that has greater control demands. The present study assessed this possibility in Spanish-English bilinguals with AD (n = 21) and cognitively normal controls (n = 30) by examining the brain correlates of dominant versus nondominant language performance on the Multilingual Naming Test (MINT), adjusting for age and education. There were no significant structural correlates of naming performance for either language in controls. In patients with AD, dominant language MINT performance was associated with cortical thickness of the entorhinal cortex and middle temporal gyrus, consistent with previous findings of temporal atrophy and related decline of naming abilities in AD. Nondominant language MINT performance, in contrast, was correlated with thickness of the left caudal anterior cingulate cortex (ACC), a central cognitive control region involved in error monitoring and task switching. The relationship between naming in the nondominant language and ACC in patients with AD but not in controls may reflect increased reliance on the ACC for nondominant language use in the face of atrophy of other control network components. The results are consistent with the possibility that the increased burden nondominant language use places on cognitive control systems compromised in AD may account for faster nondominant than dominant language decline in AD.

Shared Latent Structures Between Imaging Features and Biomarkers in Early Stages of Alzheimer's Disease: A Predictive Study

Magnetic resonance imaging (MRI) provides high resolution brain morphological information and is used as a biomarker in neurodegenerative diseases. Population studies of brain morphology often seek to identify pathological structural changes related to different diagnostic categories (e.g.: controls, mild cognitive impairment or dementia) which normally describe highly heterogeneous groups with a single categorical variable. Instead, multiple biomarkers are used as a proxy for pathology and are more powerful in capturing structural variability. Hence, using the joint modeling of brain morphology and biomarkers, we aim at describing structural changes related to any brain condition by means of few underlying processes. In this regard, we use a multivariate approach based on Projection to Latent Structures in its regression variant (PLSR) to study structural changes related to aging and AD pathology. MRI volumetric and cortical thickness measurements are used for brain morphology and cerebrospinal fluid (CSF) biomarkers (t-tau, p-tau and amyloid-beta) are used as a proxy for AD pathology. By relating both sets of measurements, PLSR finds a low-dimensional latent space describing AD pathological effects on brain structure. The proposed framework allows us to separately model aging effects on brain morphology as a confounder variable orthogonal to the pathological effect. The predictive power of the associated latent spaces (i.e., the capacity of predicting biomarker values) is assessed in a cross-validation framework.

Optimizing Machine Learning Methods to Improve Predictive Models of Alzheimer's Disease

BACKGROUND

Predicting clinical course of cognitive decline can boost clinical trials' power and improve our clinical decision-making. Machine learning (ML) algorithms are specifically designed for the purpose of prediction; however. identifying optimal features or algorithms is still a challenge.

OBJECTIVE

To investigate the accuracy of different ML methods and different features to classify cognitively normal (CN) individuals from Alzheimer's disease (AD) and to predict longitudinal outcome in participants with mild cognitive impairment (MCI).

METHODS

A total of 1,329 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were included: 424 CN, 656 MCI, and 249 AD individuals. Four feature-sets at baseline (hippocampal volume and volume of 47 cortical and subcortical regions with and without demographics and APOE4) and six machine learning methods (decision trees, support vector machines, K-nearest neighbor, ensemble linear discriminant, boosted trees, and random forests) were used to classify participants with normal cognition from participants with AD. Subsequently the model with best classification performance was used for predicting clinical outcome of MCI participants.

RESULTS

Ensemble linear discriminant models using demographics and all volumetric magnetic resonance imaging measures as feature-set showed the best performance in classification of CN versus AD participants (accuracy = 92.8%, sensitivity = 95.8%, and specificity = 88.3%). Prediction accuracy of future conversion from MCI to AD for this ensemble linear discriminant at 6, 12, 24, 36, and 48 months was 63.8% (sensitivity = 74.4, specificity = 63.1), 68.9% (sensitivity = 75.9, specificity = 67.8), 74.9% (sensitivity = 71.5, specificity = 76.3), 75.3%, (sensitivity = 65.2, specificity = 79.7), and 77.0% (sensitivity = 59.6, specificity = 86.1), respectively.

CONCLUSIONS

Machine learning models trained for classification of CN versus AD can improve our prediction ability of MCI conversion to AD.

Contourlet-based hippocampal magnetic resonance imaging texture features for multivariant classification and prediction of Alzheimer's disease

The study is aimed to assess whether the addition of contourlet-based hippocampal magnetic resonance imaging (MRI) texture features to multivariant models improves the classification of Alzheimer's disease (AD) and the prediction of mild cognitive impairment (MCI) conversion, and to evaluate whether Gaussian process (GP) and partial least squares (PLS) are feasible in developing multivariant models in this context. Clinical and MRI data of 58 patients with probable AD, 147 with MCI, and 94 normal controls (NCs) were collected. Baseline contourlet-based hippocampal MRI texture features, medical histories, symptoms, neuropsychological tests, volume-based morphometric (VBM) parameters based on MRI, and regional CMgl measurement based on fluorine-18 fluorodeoxyglucose-positron emission tomography were included to develop GP and PLS models to classify different groups of subjects. GPR1 model, which incorporated MRI texture features and was based on GPG, performed better in classifying different groups of subjects than GPR2 model, which used the same algorithm and had the same data as GPR1 except that MRI texture features were excluded. PLS model, which included the same variables as GPR1 but was based on the PLS algorithm, performed best among the three models. GPR1 accurately predicted 82.2% (51/62) of MCI convertors confirmed during the 2-year follow-up period, while this figure was 53 (85.5%) for PLS model. GPR1 and PLS models accurately predicted 58 (79.5%) vs. 61 (83.6%) of 73 patients with stable MCI, respectively. For seven patients with MCI who converted to NCs, PLS model accurately predicted all cases (100%), while GPR1 predicted six (85.7%) cases. The addition of contourlet-based MRI texture features to multivariant models can effectively improve the classification of AD and the prediction of MCI conversion to AD. Both GPR and LPS models performed well in the classification and predictive process, with the latter having significantly higher classification and predictive accuracies. Advances in knowledge: We combined contourlet-based hippocampal MRI texture features, medical histories, symptoms, neuropsychological tests, volume-based morphometric (VBM) parameters, and regional CMgl measurement to develop models using GP and PLS algorithms to classify AD patients.

Stability of graph theoretical measures in structural brain networks in Alzheimer's disease

Graph analysis has become a popular approach to study structural brain networks in neurodegenerative disorders such as Alzheimer's disease (AD). However, reported results across similar studies are often not consistent. In this paper we investigated the stability of the graph analysis measures clustering, path length, global efficiency and transitivity in a cohort of AD (N = 293) and control subjects (N = 293). More specifically, we studied the effect that group size and composition, choice of neuroanatomical atlas, and choice of cortical measure (thickness or volume) have on binary and weighted network properties and relate them to the magnitude of the differences between groups of AD and control subjects. Our results showed that specific group composition heavily influenced the network properties, particularly for groups with less than 150 subjects. Weighted measures generally required fewer subjects to stabilize and all assessed measures showed robust significant differences, consistent across atlases and cortical measures. However, all these measures were driven by the average correlation strength, which implies a limitation of capturing more complex features in weighted networks. In binary graphs, significant differences were only found in the global efficiency and transitivity measures when using cortical thickness measures to define edges. The findings were consistent across the two atlases, but no differences were found when using cortical volumes. Our findings merits future investigations of weighted brain networks and suggest that cortical thickness measures should be preferred in future AD studies if using binary networks. Further, studying cortical networks in small cohorts should be complemented by analyzing smaller, subsampled groups to reduce the risk that findings are spurious.

Longitudinal relationships among depressive symptoms, cortisol, and brain atrophy in the neocortex and the hippocampus

OBJECTIVE

Depression is associated with accelerated aging and age-related diseases. However, mechanisms underlying this relationship remain unclear. The aim of this study was to longitudinally assess the link between depressive symptoms, brain atrophy, and cortisol levels.

METHOD

Participants from the Betula prospective cohort study (mean age = 59 years, SD = 13.4 years) underwent clinical, neuropsychological and brain 3T MRI assessments at baseline and a 4-year follow-up. Cortisol levels were measured at baseline in four saliva samples. Cortical and hippocampal atrophy rates were estimated and compared between participants with and without depressive symptoms (n = 81) and correlated with cortisol levels (n = 49).

RESULTS

Atrophy in the left superior frontal gyrus and right lingual gyrus developed in parallel with depressive symptoms, and in the left temporal pole, superior temporal cortex, and supramarginal cortex after the onset of depressive symptom. Depression-related atrophy was significantly associated with elevated cortisol levels. Elevated cortisol levels were also associated with widespread prefrontal, parietal, lateral, and medial temporal atrophy.

CONCLUSION

Depressive symptoms and elevated cortisol levels are associated with atrophy of the prefrontal and limbic areas of the brain.

Classification of Alzheimer's and MCI Patients from Semantically Parcelled PET Images: A Comparison between AV45 and FDG-PET

Early identification of dementia in the early or late stages of mild cognitive impairment (MCI) is crucial for a timely diagnosis and slowing down the progression of Alzheimer's disease (AD). Positron emission tomography (PET) is considered a highly powerful diagnostic biomarker, but few approaches investigated the efficacy of focusing on localized PET-active areas for classification purposes. In this work, we propose a pipeline using learned features from semantically labelled PET images to perform group classification. A deformable multimodal PET-MRI registration method is employed to fuse an annotated MNI template to each patient-specific PET scan, generating a fully labelled volume from which 10 common regions of interest used for AD diagnosis are extracted. The method was evaluated on 660 subjects from the ADNI database, yielding a classification accuracy of 91.2% for AD versus NC when using random forests combining features from cross-sectional and follow-up exams. A considerable improvement in the early versus late MCI classification accuracy was achieved using FDG-PET compared to the AV-45 compound, yielding a 72.5% rate. The pipeline demonstrates the potential of exploiting longitudinal multiregion PET features to improve cognitive assessment.

Multimodal Classification of Mild Cognitive Impairment Based on Partial Least Squares

In recent years, increasing attention has been given to the identification of the conversion of mild cognitive impairment (MCI) to Alzheimer's disease (AD). Brain neuroimaging techniques have been widely used to support the classification or prediction of MCI. The present study combined magnetic resonance imaging (MRI), 18F-fluorodeoxyglucose PET (FDG-PET), and 18F-florbetapir PET (florbetapir-PET) to discriminate MCI converters (MCI-c, individuals with MCI who convert to AD) from MCI non-converters (MCI-nc, individuals with MCI who have not converted to AD in the follow-up period) based on the partial least squares (PLS) method. Two types of PLS models (informed PLS and agnostic PLS) were built based on 64 MCI-c and 65 MCI-nc from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. The results showed that the three-modality informed PLS model achieved better classification accuracy of 81.40%, sensitivity of 79.69%, and specificity of 83.08% compared with the single-modality model, and the three-modality agnostic PLS model also achieved better classification compared with the two-modality model. Moreover, combining the three modalities with clinical test score (ADAS-cog), the agnostic PLS model (independent data: florbetapir-PET; dependent data: FDG-PET and MRI) achieved optimal accuracy of 86.05%, sensitivity of 81.25%, and specificity of 90.77%. In addition, the comparison of PLS, support vector machine (SVM), and random forest (RF) showed greater diagnostic power of PLS. These results suggested that our multimodal PLS model has the potential to discriminate MCI-c from the MCI-nc and may therefore be helpful in the early diagnosis of AD.

Quantitative MRI Brain Studies in Mild Cognitive Impairment and Alzheimer's Disease: A Methodological Review

Classifying and predicting Alzheimer's disease (AD) in individuals with memory disorders through clinical and psychometric assessment is challenging, especially in mild cognitive impairment (MCI) subjects. Quantitative structural magnetic resonance imaging acquisition methods in combination with computer-aided diagnosis are currently being used for the assessment of AD. These acquisitions methods include voxel-based morphometry, volumetric measurements in specific regions of interest (ROIs), cortical thickness measurements, shape analysis, and texture analysis. This review evaluates the aforementioned methods in the classification of cases into one of the following three groups: normal controls, MCI, and AD subjects. Furthermore, the performance of the methods is assessed on the prediction of conversion from MCI to AD. In parallel, it is also assessed which ROIs are preferred in both classification and prognosis through the different states of the disease. Structural changes in the early stages of the disease are more pronounced in the medial temporal lobe, especially in the entorhinal cortex, whereas with disease progression, both entorhinal cortex and hippocampus offer similar discriminative power. However, for the conversion from MCI subjects to AD, entorhinal cortex provides better predictive accuracies rather than other structures, such as the hippocampus.

Factors influencing accuracy of cortical thickness in the diagnosis of Alzheimer's disease

There is great value to use of structural neuroimaging in the assessment of Alzheimer's disease (AD). However, to date, predictive value of structural imaging tend to range between 80% and 90% in accuracy and it is unclear why this is the case given that structural imaging should parallel the pathologic processes of AD. There is a possibility that clinical misdiagnosis relative to the gold standard pathologic diagnosis and/or additional brain pathologies are confounding factors contributing to reduced structural imaging classification accuracy. We examined potential factors contributing to misclassification of individuals with clinically diagnosed AD purely from cortical thickness measures. Correctly classified and incorrectly classified groups were compared across a range of demographic, biological, and neuropsychological data including cerebrospinal fluid biomarkers, amyloid imaging, white matter hyperintensity (WMH) volume, cognitive, and genetic factors. Individual subject analyses suggested that at least a portion of the control individuals misclassified as AD from structural imaging additionally harbor substantial AD biomarker pathology and risk, yet are relatively resistant to cognitive symptoms, likely due to "cognitive reserve," and therefore clinically unimpaired. In contrast, certain clinical control individuals misclassified as AD from cortical thickness had increased WMH volume relative to other controls in the sample, suggesting that vascular conditions may contribute to classification accuracy from cortical thickness measures. These results provide examples of factors that contribute to the accuracy of structural imaging in predicting a clinical diagnosis of AD, and provide important information about considerations for future work aimed at optimizing structural based diagnostic classifiers for AD.

Predictive Model of Spread of Progressive Supranuclear Palsy Using Directional Network Diffusion

Several neurodegenerative disorders including Alzheimer’s disease (AD), frontotemporal dementia (FTD), Parkinson’s disease (PD), amyotrophic lateral sclerosis, and Huntington’s disease report aggregation and transmission of pathogenic proteins between cells. The topography of these diseases in the human brain also, therefore, displays a well-characterized and stereotyped regional pattern, and a stereotyped progression over time. This is most commonly true for AD and other dementias characterized by hallmark misfolded tau or alpha-synuclein pathology. Both tau and synuclein appear to propagate within brain circuits using a shared mechanism. The most canonical synucleopathy is PD; however, much less studied is a rare disorder called progressive supranuclear palsy (PSP). The hallmark pathology and atrophy in PSP are, therefore, also highly stereotyped: initially appearing in the striatum, followed by its neighbors and connected cortical areas. In this study, we explore two mechanistic aspects hitherto unknown about the canonical network diffusion model (NDM) of spread: (a) whether the NDM can apply to other common degenerative pathologies, specifically PSP, and (b) whether the directionality of spread is important in explaining empirical data. Our results on PSP reveal two important findings: first, that PSP is amenable to the connectome-based ND modeling in the same way as previously applied to AD and FTD and, second, that the NDM fit with empirical data are significantly enhanced by using the directional rather than the non-directional form of the human connectome. Specifically, we show that both the anterograde model of transmission (some to axonal terminal) and retrograde mode explain PSP topography more accurately than non-directional transmission. Collectively, these data show that the intrinsic architecture of the structural network mediates disease spread in PSP, most likely via a process of trans-neuronal transmission. These intriguing results have several ramifications for future studies.

Looking for Alzheimer's Disease morphometric signatures using machine learning techniques

BACKGROUND

We present our results in the International challenge for automated prediction of MCI from MRI data. We evaluate the performance of MRI-based neuromorphometrics features (nMF) in the classification of Healthy Controls (HC), Mild Cognitive Impairment (MCI), converters MCI (cMCI) and Alzheimer's Disease (AD) patients.

NEW METHODS

We propose to segregate participants in three groups according to Mini Mental State Examination score (MMSEs), searching for the main nMF in each group. Then we use them to develop a Multi Classifier System (MCS). We compare the MCS against a single classifier scheme using both MMSEs+nMF and nMF only. We repeat this comparison using three state-of-the-art classification algorithms.

RESULTS

The MCS showed the best performance on both Accuracy and Area Under the Receiver Operating Curve (AUC) in comparison with single classifiers. The multiclass AUC for the MCS classification on Test Dataset were 0.83 for HC, 0.76 for cMCI, 0.65 for MCI and 0.95 for AD. Furthermore, MCS's optimum accuracy on Neurodegenerative Disease (ND) detection (AD+cMCI vs MCI+HC) was 81.0% (AUC=0.88), while the single classifiers got 71.3% (AUC=0.86) and 63.1% (AUC=0.79) for MMSEs+nMF and only nMF respectively.

COMPARISON WITH EXISTING METHOD

The proposed MCS showed a better performance than using all nMF into a single state-of-the-art classifier.

CONCLUSIONS

These findings suggest that using cognitive scoring, e.g. MMSEs, in the design of a Multi Classifier System improves performance by allowing a better selection of MRI-based features.

Alzheimer's Disease Early Diagnosis Using Manifold-Based Semi-Supervised Learning

Alzheimer's disease (AD) is currently ranked as the sixth leading cause of death in the United States and recent estimates indicate that the disorder may rank third, just behind heart disease and cancer, as a cause of death for older people. Clearly, predicting this disease in the early stages and preventing it from progressing is of great importance. The diagnosis of Alzheimer's disease (AD) requires a variety of medical tests, which leads to huge amounts of multivariate heterogeneous data. It can be difficult and exhausting to manually compare, visualize, and analyze this data due to the heterogeneous nature of medical tests; therefore, an efficient approach for accurate prediction of the condition of the brain through the classification of magnetic resonance imaging (MRI) images is greatly beneficial and yet very challenging. In this paper, a novel approach is proposed for the diagnosis of very early stages of AD through an efficient classification of brain MRI images, which uses label propagation in a manifold-based semi-supervised learning framework. We first apply voxel morphometry analysis to extract some of the most critical AD-related features of brain images from the original MRI volumes and also gray matter (GM) segmentation volumes. The features must capture the most discriminative properties that vary between a healthy and Alzheimer-affected brain. Next, we perform a principal component analysis (PCA)-based dimension reduction on the extracted features for faster yet sufficiently accurate analysis. To make the best use of the captured features, we present a hybrid manifold learning framework which embeds the feature vectors in a subspace. Next, using a small set of labeled training data, we apply a label propagation method in the created manifold space to predict the labels of the remaining images and classify them in the two groups of mild Alzheimer's and normal condition (MCI/NC). The accuracy of the classification using the proposed method is 93.86% for the Open Access Series of Imaging Studies (OASIS) database of MRI brain images, providing, compared to the best existing methods, a 3% lower error rate.

Monitoring disease progression in mild cognitive impairment: Associations between atrophy patterns, cognition, APOE and amyloid

BACKGROUND

A disease severity index (SI) for Alzheimer's disease (AD) has been proposed that summarizes MRI-derived structural measures into a single score using multivariate data analysis.

OBJECTIVES

To longitudinally evaluate the use of the SI to monitor disease progression and predict future progression to AD in mild cognitive impairment (MCI). Further, to investigate the association between longitudinal change in the SI and cognitive impairment, Apolipoprotein E (APOE) genotype as well as the levels of cerebrospinal fluid amyloid-beta 1-42 (Aβ) peptide.

METHODS

The dataset included 195 AD, 145 MCI and 228 control subjects with annual follow-up for three years, where 70 MCI subjects progressed to AD (MCI-p). For each subject the SI was generated at baseline and follow-ups using 55 regional cortical thickness and subcortical volumes measures that extracted by the FreeSurfer longitudinal stream.

RESULTS

MCI-p subjects had a faster increase of the SI over time (p < 0.001). A higher SI at baseline in MCI-p was related to progression to AD at earlier follow-ups (p < 0.001) and worse cognitive impairment (p < 0.001). AD-like MCI patients with the APOE ε4 allele and abnormal Aβ levels had a faster increase of the SI, independently (p = 0.003 and p = 0.004).

CONCLUSIONS

Longitudinal changes in the SI reflect structural brain changes and can identify MCI patients at risk of progression to AD. Disease-related brain structural changes are influenced independently by APOE genotype and amyloid pathology. The SI has the potential to be used as a sensitive tool to predict future dementia, monitor disease progression as well as an outcome measure for clinical trials.

The Evans' Index revisited: New cut-off levels for use in radiological assessment of ventricular enlargement in the elderly

BACKGROUND AND PURPOSE

Assessment of ventricular enlargement is subjective and based on the radiologist's experience. Linear indices, such as the Evans Index (EI), have been proposed as markers of ventricular volume with an EI≥0.3 indicating pathologic ventricular enlargement in any subject. However, normal range for EI measured on magnetic resonance imaging (MRI) scans are lacking in healthy elderly according to age and sex. We propose new age and sex specific cut-off values for ventricular enlargement in the elderly population.

MATERIALS AND METHODS

534 participants (53% women) aged 65-84 years; 226 patients with Alzheimer's disease (AD), and 308 healthy elderly controls (CTR) from the AddNeuroMed and ADNI studies were included. The cut-off for pathological ventricular enlargement was estimated from healthy elderly categorized into age groups of 5 years range and defined as EI 97,5 percentile (mean+2SD). Cut-off values were tested on patients with Alzheimer's disease and a small sample of patients with probable idiopathic normal pressure hydrocephalus (iNPH) to assess the sensitivity.

RESULTS

The range of the EI in healthy elderly is wide and 29% of the CTR had an EI of 0.3 or greater. The EI increases with age in both CTR and AD, and the overall EI for women were lower than for men (p<0.001). New EI cut off values for male/female: 65-69 years 0.34/0.32, 70-74 years 0.36/0.33, 75-79 years 0.37/0.34 and 80-84 years 0.37/0.36. When applying the proposed cut-offs for EI in men and women aged 65-84, they differentiated between iNPH and CTR with a sensitivity of 80% and for different age and sex categories of AD and CTR with a sensitivity and specificity of 0-27% and 91-98%, respectively.

CONCLUSION

The range of the EI measurements in healthy elderly is wide, and a cut-off value of 0.3 cannot be used to differentiate between normal and enlarged ventricles in individual cases. The proposed EI thresholds from the present study show good sensitivity for the iNPH diagnosis.

A review on neuroimaging-based classification studies and associated feature extraction methods for Alzheimer's disease and its prodromal stages

Neuroimaging has made it possible to measure pathological brain changes associated with Alzheimer's disease (AD) in vivo. Over the past decade, these measures have been increasingly integrated into imaging signatures of AD by means of classification frameworks, offering promising tools for individualized diagnosis and prognosis. We reviewed neuroimaging-based studies for AD and mild cognitive impairment classification, selected after online database searches in Google Scholar and PubMed (January, 1985-June, 2016). We categorized these studies based on the following neuroimaging modalities (and sub-categorized based on features extracted as a post-processing step from these modalities): i) structural magnetic resonance imaging [MRI] (tissue density, cortical surface, and hippocampal measurements), ii) functional MRI (functional coherence of different brain regions, and the strength of the functional connectivity), iii) diffusion tensor imaging (patterns along the white matter fibers), iv) fluorodeoxyglucose positron emission tomography (FDG-PET) (metabolic rate of cerebral glucose), and v) amyloid-PET (amyloid burden). The studies reviewed indicate that the classification frameworks formulated on the basis of these features show promise for individualized diagnosis and prediction of clinical progression. Finally, we provided a detailed account of AD classification challenges and addressed some future research directions.

Distinguishing early and late brain aging from the Alzheimer's disease spectrum: consistent morphological patterns across independent samples

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.

Subcortical Shape Changes, Hippocampal Atrophy and Cortical Thinning in Future Alzheimer's Disease Patients

Efficacy of future treatments depends on biomarkers identifying patients with mild cognitive impairment at highest risk for transitioning to Alzheimer's disease. Here, we applied recently developed analysis techniques to investigate cross-sectional differences in subcortical shape and volume alterations in patients with stable mild cognitive impairment (MCI) (n = 23, age range 59–82, 47.8% female), future converters at baseline (n = 10, age range 66–84, 90% female) and at time of conversion (age range 68–87) compared to group-wise age and gender matched healthy control subjects (n = 23, age range 61–81, 47.8% female; n = 10, age range 66–82, 80% female; n = 10, age range 68–82, 70% female). Additionally, we studied cortical thinning and global and local measures of hippocampal atrophy as known key imaging markers for Alzheimer's disease. Apart from bilateral striatal volume reductions, no morphometric alterations were found in cognitively stable patients. In contrast, we identified shape alterations in striatal and thalamic regions in future converters at baseline and at time of conversion. These shape alterations were paralleled by Alzheimer's disease like patterns of left hemispheric morphometric changes (cortical thinning in medial temporal regions, hippocampal total and subfield atrophy) in future converters at baseline with progression to similar right hemispheric alterations at time of conversion. Additionally, receiver operating characteristic curve analysis indicated that subcortical shape alterations may outperform hippocampal volume in identifying future converters at baseline. These results further confirm the key role of early cortical thinning and hippocampal atrophy in the early detection of Alzheimer's disease. But first and foremost, and by distinguishing future converters but not patients with stable cognitive abilities from cognitively normal subjects, our results support the value of early subcortical shape alterations and reduced hippocampal subfield volumes as potential markers for the early detection of Alzheimer's disease.

Detection of Conversion from Mild Cognitive Impairment to Alzheimer's Disease Using Longitudinal Brain MRI

Mild Cognitive Impairment (MCI) is an intermediate stage between healthy and Alzheimer's disease (AD). To enable early intervention it is important to identify the MCI subjects that will convert to AD in an early stage. In this paper, we provide a new method to distinguish between MCI patients that either convert to Alzheimer's Disease (MCIc) or remain stable (MCIs), using only longitudinal T1-weighted MRI. Currently, most longitudinal studies focus on volumetric comparison of a few anatomical structures, thereby ignoring more detailed development inside and outside those structures. In this study we propose to exploit the anatomical development within the entire brain, as found by a non-rigid registration approach. Specifically, this anatomical development is represented by the Stationary Velocity Field (SVF) from registration between the baseline and follow-up images. To make the SVFs comparable among subjects, we use the parallel transport method to align them in a common space. The normalized SVF together with derived features are then used to distinguish between MCIc and MCIs subjects. This novel feature space is reduced using a Kernel Principal Component Analysis method, and a linear support vector machine is used as a classifier. Extensive comparative experiments are performed to inspect the influence of several aspects of our method on classification performance, specifically the feature choice, the smoothing parameter in the registration and the use of dimensionality reduction. The optimal result from a 10-fold cross-validation using 36 month follow-up data shows competitive results: accuracy 92%, sensitivity 95%, specificity 90%, and AUC 94%. Based on the same dataset, the proposed approach outperforms two alternative ones that either depends on the baseline image only, or uses longitudinal information from larger brain areas. Good results were also obtained when scans at 6, 12, or 24 months were used for training the classifier. Besides the classification power, the proposed method can quantitatively compare brain regions that have a significant difference in development between the MCIc and MCIs groups.

Imaging biomarkers of dementia: recommended visual rating scales with teaching cases

Abstract

The diagnostic work up of dementia may benefit from structured reporting of CT and/or MRI and the use of standardised visual rating scales. We advocate a more widespread use of standardised scales as part of the workflow in clinical and research evaluation of dementia. We propose routine clinical use of rating scales for medial temporal atrophy (MTA), global cortical atrophy (GCA) and white matter hyperintensities (WMH). These scales can be used for evaluation of both CT and MRI and are efficient in routine imaging assessment in dementia, and may improve the accuracy of diagnosis. Our review provides detailed imaging examples of rating increments in each of these scales and a separate teaching file. The radiologist should relate visual ratings to the clinical assessment and other biomarkers to assist the clinician in the diagnostic decision.

Teaching points

• Clinical dementia diagnostics would benefit from structured radiological reporting.

• Standardised rating scales should be used in dementia assessment.

• It is important to relate imaging findings to the clinically suspected diagnosis.

Electronic supplementary material

The online version of this article (doi:10.1007/s13244-016-0521-6) contains supplementary material, which is available to authorized users.

Extracting patterns of morphometry distinguishing HIV associated neurodegeneration from mild cognitive impairment via group cardinality constrained classification

HIV-Associated Neurocognitive Disorder (HAND) is the most common constellation of cognitive dysfunctions in chronic HIV infected patients age 60 or older in the U.S. Only few published methods assist in distinguishing HAND from other forms of age-associated cognitive decline, such as Mild Cognitive Impairment (MCI). In this report, a data-driven, nonparameteric model to identify morphometric patterns separating HAND from MCI due to non-HIV conditions in this older age group was proposed. This model enhanced the potential for group separation by combining a smaller, longitudinal data set containing HAND samples with a larger, public data set including MCI cases. Using cross-validation, a linear model on healthy controls to harmonize the volumetric scores extracted from MRIs for demographic and acquisition differences between the two independent, disease-specific data sets was trained. Next, patterns distinguishing HAND from MCI via a group sparsity constrained logistic classifier were identified. Unlike existing approaches, our classifier directly solved the underlying minimization problem by decoupling the minimization of the logistic regression function from enforcing the group sparsity constraint. The extracted patterns consisted of eight regions that distinguished HAND from MCI on a significant level while being indifferent to differences in demographics and acquisition between the two sets. Individually selecting regions through conventional morphometric group analysis resulted in a larger number of regions that were less accurate. In conclusion, simultaneously analyzing all brain regions and time points for disease specific patterns contributed to distinguishing with high accuracy HAND-related impairment from cognitive impairment found in the HIV uninfected, MCI cohort. Hum Brain Mapp 37:4523-4538, 2016. © 2016 Wiley Periodicals, Inc.

Disrupted Network Topology in Patients with Stable and Progressive Mild Cognitive Impairment and Alzheimer's Disease

Recent findings suggest that Alzheimer's disease (AD) is a disconnection syndrome characterized by abnormalities in large-scale networks. However, the alterations that occur in network topology during the prodromal stages of AD, particularly in patients with stable mild cognitive impairment (MCI) and those that show a slow or faster progression to dementia, are still poorly understood. In this study, we used graph theory to assess the organization of structural MRI networks in stable MCI (sMCI) subjects, late MCI converters (lMCIc), early MCI converters (eMCIc), and AD patients from 2 large multicenter cohorts: ADNI and AddNeuroMed. Our findings showed an abnormal global network organization in all patient groups, as reflected by an increased path length, reduced transitivity, and increased modularity compared with controls. In addition, lMCIc, eMCIc, and AD patients showed a decreased path length and mean clustering compared with the sMCI group. At the local level, there were nodal clustering decreases mostly in AD patients, while the nodal closeness centrality detected abnormalities across all patient groups, showing overlapping changes in the hippocampi and amygdala and nonoverlapping changes in parietal, entorhinal, and orbitofrontal regions. These findings suggest that the prodromal and clinical stages of AD are associated with an abnormal network topology.

Structural Neuroimaging of the Medial Temporal Lobe in Alzheimer's Disease Clinical Trials

Atrophy in the medial temporal lobe (MTA) is being used as a criterion to support a diagnosis of Alzheimer's disease (AD). There are several structural neuroimaging approaches for quantifying MTA, including semiquantitative visual rating scales, volumetry (3D), planimetry (2D), and linear measures (1D). Current applications of structural neuroimaging in Alzheimer's disease clinical trials (ADCTs) incorporate it as a tool for improving the selection of subjects for enrollment or for stratification, for tracking disease progression, or providing evidence of target engagement for new therapeutic agents. It may also be used as a surrogate marker, providing evidence of disease-modifying effects. However, despite the widespread use of volumetric magnetic resonance imaging (MRI) in ADCTs, there are some important challenges and limitations, such as difficulties in the interpretation of results, limitations in translating results into clinical practice, and reproducibility issues, among others. Solutions to these issues may arise from other methodologies that are able to link the results of volumetric MRI from trials with conventional MRIs performed in routine clinical practice (linear or planimetric methods). Also of potential benefit are automated volumetry, using indices for comparing the relative rate of atrophy of different regions instead of absolute rates of atrophy, and combining structural neuroimaging with other biomarkers. In this review, authors present the existing structural neuroimaging approaches for MTA quantification. They then discuss solutions to the limitations of the different techniques as well as the current challenges of the field. Finally, they discuss how the current advances in AD neuroimaging can help AD diagnosis.

Reproducibility of hippocampal atrophy rates measured with manual, FreeSurfer, AdaBoost, FSL/FIRST and the MAPS-HBSI methods in Alzheimer's disease

The purpose of this study is to assess the reproducibility of hippocampal atrophy rate measurements of commonly used fully-automated algorithms in Alzheimer disease (AD). The reproducibility of hippocampal atrophy rate for FSL/FIRST, AdaBoost, FreeSurfer, MAPS independently and MAPS combined with the boundary shift integral (MAPS-HBSI) were calculated. Back-to-back (BTB) 3D T1-weighted MPRAGE MRI from the Alzheimer's Disease Neuroimaging Initiative (ADNI1) study at baseline and year one were used. Analysis on 3 groups of subjects was performed - 562 subjects at 1.5T, a 75 subject group that also had manual segmentation and 111 subjects at 3T. A simple and novel statistical test based on the binomial distribution was used that handled outlying data points robustly. Median hippocampal atrophy rates were -1.1%/year for healthy controls, -3.0%/year for mildly cognitively impaired and -5.1%/year for AD subjects. The best reproducibility was observed for MAPS-HBSI (1.3%), while the other methods tested had reproducibilities at least 50% higher at 1.5T and 3T which was statistically significant. For a clinical trial, MAPS-HBSI should require less than half the subjects of the other methods tested. All methods had good accuracy versus manual segmentation. The MAPS-HBSI method has substantially better reproducibility than the other methods considered.

The effect of increased genetic risk for Alzheimer's disease on hippocampal and amygdala volume

Reduction in hippocampal and amygdala volume measured via structural magnetic resonance imaging is an early marker of Alzheimer's disease (AD). Whether genetic risk factors for AD exert an effect on these subcortical structures independent of clinical status has not been fully investigated. We examine whether increased genetic risk for AD influences hippocampal and amygdala volumes in case-control and population cohorts at different ages, in 1674 older (aged >53 years; 17% AD, 39% mild cognitive impairment [MCI]) and 467 young (16-30 years) adults. An AD polygenic risk score combining common risk variants excluding apolipoprotein E (APOE), and a single nucleotide polymorphism in TREM2, were both associated with reduced hippocampal volume in healthy older adults and those with MCI. APOE ε4 was associated with hippocampal and amygdala volume in those with AD and MCI but was not associated in healthy older adults. No associations were found in young adults. Genetic risk for AD affects the hippocampus before the clinical symptoms of AD, reflecting a neurodegenerative effect before clinical manifestations in older adults.

Manual Planimetry of the Medial Temporal Lobe Versus Automated Volumetry of the Hippocampus in the Diagnosis of Alzheimer's Disease

Introduction:

Though a disproportionate rate of atrophy in the medial temporal lobe (MTA) represents a reliable marker of Alzheimer’s disease (AD) pathology, measurement of the MTA is not currently widely used in daily clinical practice. This is mainly because the methods available to date are sophisticated and difficult to implement in clinical practice (volumetric methods), are poorly explored (linear and planimetric methods), or lack objectivity (visual rating). Here, we aimed to compare the results of a manual planimetric measure (the yearly rate of absolute atrophy of the medial temporal lobe, 2D-yrA-MTL) with the results of an automated volumetric measure (the yearly rate of atrophy of the hippocampus, 3D-yrA-H).

Methods:

A series of 1.5T MRI studies on 290 subjects in the age range of 65–85 years, including patients with AD (n = 100), mild cognitive impairment (MCI) (n = 100), and matched controls (n = 90) from the AddNeuroMed study, were examined by two independent subgroups of researchers: one in charge of volumetric measures and the other in charge of planimetric measures.

Results:

The means of both methods were significantly different between AD and the other two diagnostic groups. In the differential diagnosis of AD against controls, 3D-yrA-H performed significantly better than 2D-yrA-MTL while differences were not statistically significant in the differential diagnosis of AD against MCI.

Conclusion:

Automated volumetry of the hippocampus is superior to manual planimetry of the MTL in the diagnosis of AD. Nevertheless, the 2D-yrAMTL is a simpler method that could be easily implemented in clinical practice when volumetry is not available. 

Measuring Global Brain Atrophy with the Brain Volume/Cerebrospinal Fluid Index: Normative Values, Cut-Offs and Clinical Associations

BACKGROUND

Global brain atrophy is present in normal aging and different neurodegenerative disorders such as Alzheimer's disease (AD) and is becoming widely used to monitor disease progression.

SUMMARY

The brain volume/cerebrospinal fluid index (BV/CSF index) is validated in this study as a measurement of global brain atrophy. We tested the ability of the BV/CSF index to detect global brain atrophy, investigated the influence of confounders, provided normative values and cut-offs for mild, moderate and severe brain atrophy, and studied associations with different outcome variables. A total of 1,009 individuals were included [324 healthy controls, 408 patients with mild cognitive impairment (MCI) and 277 patients with AD]. Magnetic resonance images were segmented using FreeSurfer, and the BV/CSF index was calculated and studied both cross-sectionally and longitudinally (1-year follow-up). Both AD patients and MCI patients who progressed to AD showed greater global brain atrophy compared to stable MCI patients and controls. Atrophy was associated with older age, larger intracranial volume, less education and presence of the ApoE ε4 allele. Significant correlations were found with clinical variables, CSF biomarkers and several cognitive tests.

KEY MESSAGES

The BV/CSF index may be useful for staging individuals according to the degree of global brain atrophy, and for monitoring disease progression. It also shows potential for predicting clinical changes and for being used in the clinical routine.

Detection of Alzheimer's disease signature in MR images seven years before conversion to dementia: Toward an early individual prognosis

Finding very early biomarkers of Alzheimer's Disease (AD) to aid in individual prognosis is of major interest to accelerate the development of new therapies. Among the potential biomarkers, neurodegeneration measurements from MRI are considered as good candidates but have so far not been effective at the early stages of the pathology. Our objective is to investigate the efficiency of a new MR-based hippocampal grading score to detect incident dementia in cognitively intact patients. This new score is based on a pattern recognition strategy, providing a grading measure that reflects the similarity of the anatomical patterns of the subject under study with dataset composed of healthy subjects and patients with AD. Hippocampal grading was evaluated on subjects from the Three-City cohort, with a followup period of 12 years. Experiments demonstrate that hippocampal grading yields prediction accuracy up to 72.5% (P < 0.0001) 7 years before conversion to AD, better than both hippocampal volume (58.1%, P = 0.04) and MMSE score (56.9%, P = 0.08). The area under the ROC curve (AUC) supports the efficiency of imaging biomarkers with a gain of 8.4 percentage points for hippocampal grade (73.0%) over hippocampal volume (64.6%). Adaptation of the proposed framework to clinical score estimation is also presented. Compared with previous studies investigating new biomarkers for AD prediction over much shorter periods, the very long followup of the Three-City cohort demonstrates the important clinical potential of the proposed imaging biomarker. The high accuracy obtained with this new imaging biomarker paves the way for computer-based prognostic aides to help the clinician identify cognitively intact subjects that are at high risk to develop AD.