Atypical brain structure and function in young adults exposed to disaster-related prenatal maternal stress: Project Ice Storm.

Studies have shown that prenatal maternal stress (PNMS) affects brain structure and function in childhood. However, less research has examined whether PNMS effects on brain structure and function extend to young adulthood. We recruited women who were pregnant during or within 3 months following the 1998 Quebec ice storm, assessed their PNMS, and prospectively followed-up their children. T1-weighted magnetic resonance imaging (MRI) and resting-state functional MRI were obtained from 19-year-old young adults with (n = 39) and without (n = 65) prenatal exposure to the ice storm. We examined between-group differences in gray matter volume (GMV), surface area (SA), and cortical thickness (CT). We used the brain regions showing between-group GMV differences as seeds to compare between-group functional connectivity. Within the Ice Storm group, we examined (1) associations between PNMS and the atypical GMV, SA, CT, and functional connectivity, and (2) moderation by timing of exposure. Primarily, we found that, compared to Controls, the Ice Storm youth had larger GMV and higher functional connectivity of the anterior cingulate cortex, the precuneus, the left occipital pole, and the right hippocampus; they also had larger CT, but not SA, of the left occipital pole. Within the Ice Storm group, maternal subjective distress during preconception and mid-to-late pregnancy was associated with atypical left occipital pole CT. These results suggest the long-lasting impact of disaster-related PNMS on child brain structure and functional connectivity. Our study also indicates timing-specific effects of the subjective aspect of PNMS on occipital thickness.

Neural correlates of disaster-related prenatal maternal stress in young adults from Project Ice Storm: Focus on amygdala, hippocampus, and prefrontal cortex.

Background: Studies have shown that prenatal maternal stress alters volumes of the amygdala and hippocampus, and alters functional connectivity between the amygdala and prefrontal cortex. However, it remains unclear whether prenatal maternal stress (PNMS) affects volumes and functional connectivity of these structures at their subdivision levels. Methods: T1-weighted MRI and resting-state functional MRI were obtained from 19-year-old young adult offspring with (n = 39, 18 male) and without (n = 65, 30 male) exposure to PNMS deriving from the 1998 ice storm. Volumes of amygdala nuclei, hippocampal subfields and prefrontal subregions were computed, and seed-to-seed functional connectivity analyses were conducted. Results: Compared to controls, young adult offspring exposed to disaster-related PNMS had larger volumes of bilateral whole amygdala, driven by the lateral, basal, central, medial, cortical, accessory basal nuclei, and corticoamygdaloid transition; larger volumes of bilateral whole hippocampus, driven by the CA1, HATA, molecular layer, fissure, tail, CA3, CA4, and DG; and larger volume of the prefrontal cortex, driven by the left superior frontal. Inversely, young adult offspring exposed to disaster-related PNMS had lower functional connectivity between the whole amygdala and the prefrontal cortex (driven by bilateral frontal poles, the left superior frontal and left caudal middle frontal); and lower functional connectivity between the hippocampal tail and the prefrontal cortex (driven by the left lateral orbitofrontal). Conclusion: These results suggest the possibility that effects of disaster-related PNMS on structure and function of subdivisions of offspring amygdala, hippocampus and prefrontal cortex could persist into young adulthood.

Intrinsic connectivity of the human brain provides scaffold for tau aggregation in clinical variants of Alzheimer's disease.

Alzheimer's disease (AD) phenotypes might result from differences in selective vulnerability. Evidence from preclinical models suggests that tau pathology has cell-to-cell propagation properties. Therefore, here, we tested the cell-to-cell propagation framework in the amnestic, visuospatial, language, and behavioral/dysexecutive phenotypes of AD. We report that each AD phenotype is associated with a distinct network-specific pattern of tau aggregation, where tau aggregation is concentrated in brain network hubs. In all AD phenotypes, regional tau load could be predicted by connectivity patterns of the human brain. Furthermore, regions with greater connectivity displayed similar rates of longitudinal tau accumulation in an independent cohort. Connectivity-based tau deposition was not restricted to a specific vulnerable network but was rather a general property of brain organization, linking selective vulnerability and transneuronal spreading models of neurodegeneration. Together, this study indicates that intrinsic brain connectivity provides a framework for tau aggregation across diverse phenotypic manifestations of AD.

18F-MK-6240 PET for early and late detection of neurofibrillary tangles.

Braak stages of tau neurofibrillary tangle accumulation have been incorporated in the criteria for the neuropathological diagnosis of Alzheimer's disease. It is expected that Braak staging using brain imaging can stratify living individuals according to their individual patterns of tau deposition, which may prove crucial for clinical trials and practice. However, previous studies using the first-generation tau PET agents have shown a low sensitivity to detect tau pathology in areas corresponding to early Braak histopathological stages (∼20% of cognitively unimpaired elderly with tau deposition in regions corresponding to Braak I-II), in contrast to ∼80-90% reported in post-mortem cohorts. Here, we tested whether the novel high affinity tau tangles tracer 18F-MK-6240 can better identify individuals in the early stages of tau accumulation. To this end, we studied 301 individuals (30 cognitively unimpaired young, 138 cognitively unimpaired elderly, 67 with mild cognitive impairment, 54 with Alzheimer's disease dementia, and 12 with frontotemporal dementia) with amyloid-ß 18F-NAV4694, tau 18F-MK-6240, MRI, and clinical assessments. 18F-MK-6240 standardized uptake value ratio images were acquired at 90-110 min after the tracer injection. 18F-MK-6240 discriminated Alzheimer's disease dementia from mild cognitive impairment and frontotemporal dementia with high accuracy (∼85-100%). 18F-MK-6240 recapitulated topographical patterns consistent with the six hierarchical stages proposed by Braak in 98% of our population. Cognition and amyloid-ß status explained most of the Braak stages variance (P < 0.0001, R2 = 0.75). No single region of interest standardized uptake value ratio accurately segregated individuals into the six topographic Braak stages. Sixty-eight per cent of the cognitively unimpaired elderly amyloid-ß-positive and 37% of the cognitively unimpaired elderly amyloid-ß-negative subjects displayed tau deposition, at least in the transentorhinal cortex (Braak I). Tau deposition solely in the transentorhinal cortex was associated with an elevated prevalence of amyloid-ß, neurodegeneration, and cognitive impairment (P < 0.0001). 18F-MK-6240 deposition in regions corresponding to Braak IV-VI was associated with the highest prevalence of neurodegeneration, whereas in Braak V-VI regions with the highest prevalence of cognitive impairment. Our results suggest that the hierarchical six-stage Braak model using 18F-MK-6240 imaging provides an index of early and late tau accumulation as well as disease stage in preclinical and symptomatic individuals. Tau PET Braak staging using high affinity tracers has the potential to be incorporated in the diagnosis of living patients with Alzheimer's disease in the near future.

Mild behavioral impairment is associated with ß-amyloid but not tau or neurodegeneration in cognitively intact elderly individuals.

INTRODUCTION: Mild behavioral impairment (MBI) is characterized by the emergence of neuropsychiatric symptoms in elderly persons. Here, we examine the associations between MBI and Alzheimer's disease (AD) biomarkers in asymptomatic elderly individuals. METHODS: Ninety-six cognitively normal elderly individuals underwent MRI, [18 F]AZD4694 ß-amyloid-PET, and [18 F]MK6240 tau-PET. MBI was assessed using the MBI Checklist (MBI-C). Pearson's correlations and voxel-based regressions were used to evaluate the relationship between MBI-C score and [18 F]AZD4694 retention, [18 F]MK6240 retention, and gray matter (GM) volume. RESULTS: Pearson correlations revealed a positive relationship between MBI-C score and global and striatal [18 F]AZD4694 standardized uptake value ratios (SUVRs). Voxel-based regression analyses revealed a positive correlation between MBI-C score and [18 F]AZD4694 retention. No significant correlations were found between MBI-C score and [18 F]MK6240 retention or GM volume. CONCLUSION: We demonstrate for the first time a link between MBI and early AD pathology in a cognitively intact elderly population, supporting the use of the MBI-C as a metric to enhance clinical trial enrolment.

3D-Deep Learning Based Automatic Diagnosis of Alzheimer's Disease with Joint MMSE Prediction Using Resting-State fMRI.

We performed this research to 1) evaluate a novel deep learning method for the diagnosis of Alzheimer's disease (AD) and 2) jointly predict the Mini Mental State Examination (MMSE) scores of South Korean patients with AD. Using resting-state functional Magnetic Resonance Imaging (rs-fMRI) scans of 331 participants, we obtained functional 3-dimensional (3-D) independent component spatial maps for use as features in classification and regression tasks. A 3-D convolutional neural network (CNN) architecture was developed for the classification task. MMSE scores were predicted using: linear least square regression (LLSR), support vector regression, bagging-based ensemble regression, and tree regression with group independent component analysis (gICA) features. To improve MMSE regression performance, we applied feature optimization methods including least absolute shrinkage and selection operator and support vector machine-based recursive feature elimination (SVM-RFE). The mean balanced test accuracy was 85.27% for the classification of AD versus healthy controls. The medial visual, default mode, dorsal attention, executive, and auditory related networks were mainly associated with AD. The maximum clinical MMSE score prediction accuracy with the LLSR method applied on gICA combined with SVM-RFE features had the lowest root mean square error (3.27 ± 0.58) and the highest R2 value (0.63 ± 0.02). Classification of AD and healthy controls can be successfully achieved using only rs-fMRI and MMSE scores can be accurately predicted using functional independent component features. In the absence of trained clinicians, AD disease status and clinical MMSE scores can be jointly predicted using 3-D deep learning and regression learning approaches with rs-fMRI data.

Association of Apolipoprotein E ε4 With Medial Temporal Tau Independent of Amyloid-ß.

Importance: Apolipoprotein E ε4 (APOEε4) is the single most important genetic risk factor for Alzheimer disease. While APOEε4 is associated with increased amyloid-ß burden, its association with cerebral tau pathology has been controversial. Objective: To determine whether APOEε4 is associated with medial temporal tau pathology independently of amyloid-ß, sex, clinical status, and age. Design, Setting, and Participants: This is a study of 2 cross-sectional cohorts of volunteers who were cognitively normal, had mild cognitive impairment (MCI), or had Alzheimer disease dementia: the Translational Biomarkers in Aging and Dementia (TRIAD) study (data collected between October 2017 and July 2019) and the Alzheimer's Disease Neuroimaging Initiative (ADNI) (collected between November 2015 and June 2019). The first cohort (TRIAD) comprised cognitively normal elderly participants (n = 124), participants with MCI (n = 50), and participants with Alzheimer disease (n = 50) who underwent tau positron emission tomography (PET) with fluorine 18-labeled MK6240 and amyloid-ß PET with [18F]AZD4694. The second sample (ADNI) was composed of cognitively normal elderly participants (n = 157), participants with MCI (n = 83), and participants with Alzheimer disease (n = 25) who underwent tau PET with [18F]flortaucipir and amyloid-ß PET with [18F]florbetapir. Exclusion criteria were a history of other neurological disorders, stroke, or head trauma. There were 489 eligible participants, selected based on availability of amyloid-PET, tau-PET, magnetic resonance imaging, and genotyping for APOEε4. Forty-five young adults (<30 years) from the TRIAD cohort were not selected for this study. Main Outcomes and Measures: A main association between APOEε4 and tau-PET standardized uptake value ratio, correcting for age, sex, clinical status, and neocortical amyloid-PET standardized uptake value ratio. Results: The mean (SD) age of the 489 participants was 70.5 (7.1) years; 171 were APOEε4 carriers (34.9%), and 230 of 489 were men. In both cohorts, APOEε4 was associated in increased tau-PET uptake in the entorhinal cortex and hippocampus independently of amyloid-ß, sex, age, and clinical status after multiple comparisons correction (TRIAD: ß = 0.33; 95% CI, 0.19-0.49; ADNI: ß = 0.13; 95% CI, 0.08-0.19; P < .001). Conclusions and Relevance: Our results indicate that the elevated risk of developing dementia conferred by APOEε4 genotype involves mechanisms associated with both amyloid-ß and tau aggregation. These results contribute to an evolving framework in which APOEε4 has deleterious consequences in Alzheimer disease beyond its link with amyloid-ß and suggest APOEε4 as a potential target for future disease-modifying therapeutic trials targeting tau pathology.

3D-CNN based discrimination of schizophrenia using resting-state fMRI.

MOTIVATION: This study reports a framework to discriminate patients with schizophrenia and normal healthy control subjects, based on magnetic resonance imaging (MRI) of the brain. Resting-state functional MRI data from a total of 144 subjects (72 patients with schizophrenia and 72 healthy controls) was obtained from a publicly available dataset using a three-dimensional convolution neural network 3D-CNN based deep learning classification framework and ICA based features. RESULTS: We achieved 98.09 ± 1.01% ten-fold cross-validated classification accuracy with a p-value < 0.001 and an area under the curve (AUC) of 0.9982 ± 0.015. In addition, differences in functional connectivity between the two groups were statistically analyzed across multiple resting-state networks. The disconnection between the visual and frontal network was prominent in patients, while they showed higher connectivity between the default mode network and other task-positive/ cerebellar networks. These ICA functional network maps served as highly discriminative three-dimensional imaging features for the discrimination of schizophrenia in this study. CONCLUSION: Due to the very high AUC, this research with more validation on the cross diagnosis and publicly available dataset, may be translated in future as an adjunct tool to assist clinicians in the initial screening of schizophrenia.

Evaluation of Functional Decline in Alzheimer's Dementia Using 3D Deep Learning and Group ICA for rs-fMRI Measurements.

Purpose: To perform automatic assessment of dementia severity using a deep learning framework applied to resting-state functional magnetic resonance imaging (rs-fMRI) data. Method: We divided 133 Alzheimer's disease (AD) patients with clinical dementia rating (CDR) scores from 0.5 to 3 into two groups based on dementia severity; the groups with very mild/mild (CDR: 0.5-1) and moderate to severe (CDR: 2-3) dementia consisted of 77 and 56 subjects, respectively. We used rs-fMRI to extract functional connectivity features, calculated using independent component analysis (ICA), and performed automated severity classification with three-dimensional convolutional neural networks (3D-CNNs) based on deep learning. Results: The mean balanced classification accuracy was 0.923 ± 0.042 (p < 0.001) with a specificity of 0.946 ± 0.019 and sensitivity of 0.896 ± 0.077. The rs-fMRI data indicated that the medial frontal, sensorimotor, executive control, dorsal attention, and visual related networks mainly correlated with dementia severity. Conclusions: Our CDR-based novel classification using rs-fMRI is an acceptable objective severity indicator. In the absence of trained neuropsychologists, dementia severity can be objectively and accurately classified using a 3D-deep learning framework with rs-fMRI independent components.

Hybrid multivariate pattern analysis combined with extreme learning machine for Alzheimer's dementia diagnosis using multi-measure rs-fMRI spatial patterns.

BACKGROUND: Early diagnosis of Alzheimer's disease (AD) and Mild Cognitive Impairment (MCI) is essential for timely treatment. Machine learning and multivariate pattern analysis (MVPA) for the diagnosis of brain disorders are explicitly attracting attention in the neuroimaging community. In this paper, we propose a voxel-wise discriminative framework applied to multi-measure resting-state fMRI (rs-fMRI) that integrates hybrid MVPA and extreme learning machine (ELM) for the automated discrimination of AD and MCI from the cognitive normal (CN) state. MATERIALS AND METHODS: We used two rs-fMRI cohorts: the public Alzheimer's disease Neuroimaging Initiative database (ADNI2) and an in-house Alzheimer's disease cohort from South Korea, both including individuals with AD, MCI, and normal controls. After extracting three-dimensional (3-D) patterns measuring regional coherence and functional connectivity during the resting state, we performed univariate statistical t-tests to generate a 3-D mask that retained only voxels showing significant changes. Given the initial univariate features, to enhance discriminative patterns, we implemented MVPA feature reduction using support vector machine-recursive feature elimination (SVM-RFE), and least absolute shrinkage and selection operator (LASSO), in combination with the univariate t-test. Classifications were performed by an ELM, and its efficiency was compared to linear and nonlinear (radial basis function) SVMs. RESULTS: The maximal accuracies achieved by the method in the ADNI2 cohort were 98.86% (p<0.001) and 98.57% (p<0.001) for AD and MCI vs. CN, respectively. In the in-house cohort, the same accuracies were 98.70% (p<0.001) and 94.16% (p<0.001). CONCLUSION: From a clinical perspective, combining extreme learning machine and hybrid MVPA applied on concatenations of multiple rs-fMRI biomarkers can potentially assist the clinicians in AD and MCI diagnosis.

Multiclass Classification of Word Imagination Speech with Hybrid Connectivity Features.

OBJECTIVE: In this study, electroencephalography data of imagined words were classified using four different feature extraction approaches. Eight subjects were recruited for the recording of imagination with five different words, namely; 'go', 'back', 'left', 'right', and 'stop'.

EEG classification for motor imagery BCI using phase-only features extracted by independent component analysis.

The accurate classification of the electroencephalography (EEG) signals is the most important task towards the development of a reliable motor imagery brain-computer interface (MI-BCI) system. In this study, we utilized a publically available BCI Competition-IV 2008 dataset IIa. This study address to the binary classification problem of the motor imagery EEG data by using a sigmoid activation function-based extreme learning machines (ELM). We proposed a novel method of extracting the features from the EEG signals by first applying the independent component analysis (ICA) on the time series data and transforming the ICA time series data into Fourier domain and then extract the phase information from the Fourier spectrum. This phase information was further used to calculate the maximized cross-correlation connectivity matrix. The upper diagonal of this matrix was then vectorized and it serves as the basic feature for the ELM classification framework. By using the phase-only features we achieved 97.80% (p <; 0.0022) nested cross-validated classification accuracy. In addition, this process is relatively computationally inexpensive. Thus, it can be an excellent candidate for the motor imagery BCI applications.

Multimodal Discrimination of Schizophrenia Using Hybrid Weighted Feature Concatenation of Brain Functional Connectivity and Anatomical Features with an Extreme Learning Machine.

Multimodal features of structural and functional magnetic resonance imaging (MRI) of the human brain can assist in the diagnosis of schizophrenia. We performed a classification study on age, sex, and handedness-matched subjects. The dataset we used is publicly available from the Center for Biomedical Research Excellence (COBRE) and it consists of two groups: patients with schizophrenia and healthy controls. We performed an independent component analysis and calculated global averaged functional connectivity-based features from the resting-state functional MRI data for all the cortical and subcortical anatomical parcellation. Cortical thickness along with standard deviation, surface area, volume, curvature, white matter volume, and intensity measures from the cortical parcellation, as well as volume and intensity from sub-cortical parcellation and overall volume of cortex features were extracted from the structural MRI data. A novel hybrid weighted feature concatenation method was used to acquire maximal 99.29% (P < 0.0001) accuracy which preserves high discriminatory power through the weight of the individual feature type. The classification was performed by an extreme learning machine, and its efficiency was compared to linear and non-linear (radial basis function) support vector machines, linear discriminant analysis, and random forest bagged tree ensemble algorithms. This article reports the predictive accuracy of both unimodal and multimodal features after 10-by-10-fold nested cross-validation. A permutation test followed the classification experiment to assess the statistical significance of the classification results. It was concluded that, from a clinical perspective, this feature concatenation approach may assist the clinicians in schizophrenia diagnosis.

Corrigendum: Multi-modal, Multi-measure, and Multi-class Discrimination of ADHD with Hierarchical Feature Extraction and Extreme Learning Machine Using Structural and Functional Brain MRI.

[This corrects the article on p. 157 in vol. 11, PMID: 28420972.].

Multi-modal, Multi-measure, and Multi-class Discrimination of ADHD with Hierarchical Feature Extraction and Extreme Learning Machine Using Structural and Functional Brain MRI.

Structural and functional MRI unveil many hidden properties of the human brain. We performed this multi-class classification study on selected subjects from the publically available attention deficit hyperactivity disorder ADHD-200 dataset of patients and healthy children. The dataset has three groups, namely, ADHD inattentive, ADHD combined, and typically developing. We calculated the global averaged functional connectivity maps across the whole cortex to extract anatomical atlas parcellation based features from the resting-state fMRI (rs-fMRI) data and cortical parcellation based features from the structural MRI (sMRI) data. In addition, the preprocessed image volumes from both of these modalities followed an ANOVA analysis separately using all the voxels. This study utilized the average measure from the most significant regions acquired from ANOVA as features for classification in addition to the multi-modal and multi-measure features of structural and functional MRI data. We extracted most discriminative features by hierarchical sparse feature elimination and selection algorithm. These features include cortical thickness, image intensity, volume, cortical thickness standard deviation, surface area, and ANOVA based features respectively. An extreme learning machine performed both the binary and multi-class classifications in comparison with support vector machines. This article reports prediction accuracy of both unimodal and multi-modal features from test data. We achieved 76.190% (p < 0.0001) classification accuracy in multi-class settings as well as 92.857% (p < 0.0001) classification accuracy in binary settings. In addition, we found ANOVA-based significant regions of the brain that also play a vital role in the classification of ADHD. Thus, from a clinical perspective, this multi-modal group analysis approach with multi-measure features may improve the accuracy of the ADHD differential diagnosis.

Multiclass Classification for the Differential Diagnosis on the ADHD Subtypes Using Recursive Feature Elimination and Hierarchical Extreme Learning Machine: Structural MRI Study.

The classification of neuroimaging data for the diagnosis of certain brain diseases is one of the main research goals of the neuroscience and clinical communities. In this study, we performed multiclass classification using a hierarchical extreme learning machine (H-ELM) classifier. We compared the performance of this classifier with that of a support vector machine (SVM) and basic extreme learning machine (ELM) for cortical MRI data from attention deficit/hyperactivity disorder (ADHD) patients. We used 159 structural MRI images of children from the publicly available ADHD-200 MRI dataset. The data consisted of three types, namely, typically developing (TDC), ADHD-inattentive (ADHD-I), and ADHD-combined (ADHD-C). We carried out feature selection by using standard SVM-based recursive feature elimination (RFE-SVM) that enabled us to achieve good classification accuracy (60.78%). In this study, we found the RFE-SVM feature selection approach in combination with H-ELM to effectively enable the acquisition of high multiclass classification accuracy rates for structural neuroimaging data. In addition, we found that the most important features for classification were the surface area of the superior frontal lobe, and the cortical thickness, volume, and mean surface area of the whole cortex.

Classification of ADHD subgroup with recursive feature elimination for structural brain MRI.

This article reports the binary classification results of ADHD patients among three subgroups by using ADHD-200 dataset. We have proposed a modified feature selection approach using standard RFE-SVM model. Our results show the significance of the proposed method by making a comparison of J-statistics, F1-score and classification accuracy based on the feature selection from the original RFE-SVM vs. the proposed modification of RFE-SVM. In addition, we have also compared the number of features in each setting to achieve the highest accuracy. After ten-fold cross-validation, we have achieved 84.17% accuracy using a linear SVM classifier. Moreover, we have found significant anatomical regions that can serve as a potential biomarker for the ADHD subgroups classification.