The Alzheimer's Disease Neuroimaging Initiative 3: Continued innovation for clinical trial improvement

ATRI EDC: a novel cloud-native remote data capture system for large multicenter Alzheimer's disease and Alzheimer's disease-related dementias clinical trails

OBJECTIVE

The Alzheimer's Therapeutic Research Institute (ATRI) developed a novel clinical data management system, the ATRI electronic data capture system (ATRI EDC), to address the complex regulatory, operational, and data requirements that arise in the conduct of multicenter Alzheimer's disease and Alzheimer's disease-related dementias (AD/ADRDs) clinical trials. We describe the system, its utility, and the broader implications for the field of clinical trials and clinical research informatics.

MATERIALS AND METHODS

The ATRI EDC system was developed, tested, and validated using community-based agile software development methods and cloud-native single-page application design principles. It offers an increasing number of application modules, supports a high degree of study-specific configuration, and empowers study teams to effectively communicate and collaborate on the accurate and timely completion of study activities.

RESULTS

To date, the ATRI EDC system supports 10 clinical studies, collecting study data for 4596 participants. Three case descriptions further illustrate how the system's capabilities support diverse study-specific requirements.

DISCUSSION

The ATRI EDC system has several advantages: its modular capabilities can accommodate rapidly evolving research designs and technologies; its community-based agile development approach and community-friendly licensing model encourage collaboration per the principles of open science; finally, with continued development and community building efforts, the system has the potential to facilitate the effective conduct of clinical studies beyond the field of AD/ADRD.

CONCLUSION

By effectively addressing the requirements of multicenter AD/ADRD studies, the ATRI EDC system supports ATRI's scientific mission of rigorously testing new AD/ADRD therapies and facilitating the effective conduct of multicenter clinical studies.

Using the Alzheimer's Disease Neuroimaging Initiative to improve early detection, diagnosis, and treatment of Alzheimer's disease

INTRODUCTION

The Alzheimer's Disease Neuroimaging Initiative (ADNI) has accumulated 15 years of clinical, neuroimaging, cognitive, biofluid biomarker and genetic data, and biofluid samples available to researchers, resulting in more than 3500 publications. This review covers studies from 2018 to 2020.

METHODS

We identified 1442 publications using ADNI data by conventional search methods and selected impactful studies for inclusion.

RESULTS

Disease progression studies supported pivotal roles for regional amyloid beta (Aβ) and tau deposition, and identified underlying genetic contributions to Alzheimer's disease (AD). Vascular disease, immune response, inflammation, resilience, and sex modulated disease course. Biologically coherent subgroups were identified at all clinical stages. Practical algorithms and methodological changes improved determination of Aβ status. Plasma Aβ, phosphorylated tau181, and neurofilament light were promising noninvasive biomarkers. Prognostic and diagnostic models were externally validated in ADNI but studies are limited by lack of ethnocultural cohort diversity.

DISCUSSION

ADNI has had a profound impact in improving clinical trials for AD.

Deep learning-based image reconstruction and post-processing methods in positron emission tomography for low-dose imaging and resolution enhancement

Image processing plays a crucial role in maximising diagnostic quality of positron emission tomography (PET) images. Recently, deep learning methods developed across many fields have shown tremendous potential when applied to medical image enhancement, resulting in a rich and rapidly advancing literature surrounding this subject. This review encapsulates methods for integrating deep learning into PET image reconstruction and post-processing for low-dose imaging and resolution enhancement. A brief introduction to conventional image processing techniques in PET is firstly presented. We then review methods which integrate deep learning into the image reconstruction framework as either deep learning-based regularisation or as a fully data-driven mapping from measured signal to images. Deep learning-based post-processing methods for low-dose imaging, temporal resolution enhancement and spatial resolution enhancement are also reviewed. Finally, the challenges associated with applying deep learning to enhance PET images in the clinical setting are discussed and future research directions to address these challenges are presented.

Assessing and validating reliable change across ADNI protocols

OBJECTIVE

Reliable change methods can aid in determining whether changes in cognitive performance over time are meaningful. The current study sought to develop and cross-validate 12-month standardized regression-based (SRB) equations for the neuropsychological measures commonly administered in the Alzheimer's Disease Neuroimaging Initiative (ADNI) longitudinal study.

METHOD

Prediction algorithms were developed using baseline score, retest interval, the presence/absence of a 6-month evaluation, age, education, sex, and ethnicity in two different samples (n = 192 each) of robustly cognitively intact community-dwelling older adults from ADNI - matched for demographic and testing factors. The developed formulae for each sample were then applied to one of the samples to determine goodness-of-fit and appropriateness of combining samples for a single set of SRB equations.

RESULTS

Minimal differences were seen between Observed 12-month and Predicted 12-month scores on most neuropsychological tests from ADNI, and when compared across samples the resultant Predicted 12-month scores were highly correlated. As a result, samples were combined and SRB prediction equations were successfully developed for each of the measures.

CONCLUSIONS

Establishing cross-validation for these SRB prediction equations provides initial support of their use to detect meaningful change in the ADNI sample, and provides the basis for future research with clinical samples to evaluate potential clinical utility. While some caution should be considered for measuring true cognitive change over time - particularly in clinical samples - when using these prediction equations given the relatively lower coefficients of stability observed, use of these SRBs reflects an improvement over current practice in ADNI.

Increased retention of tau PET ligand [18F]-AV1451 in Alzheimer's Disease Psychosis

Psychosis in Alzheimer's disease (AD) represents a distinct disease subtype with a more rapid progression of illness evidenced by an increased velocity of cognitive decline and a hastened mortality. Previous biomarker and post-mortem studies have implicated tau neuropathology as a possible mediator of the accelerated decline in AD psychosis. Tau positron emission tomography (PET) neuroimaging provides the opportunity to evaluate tau pathology in-vivo, so that clinical symptomatology can be correlated with disease pathology. [¹⁸F]-AV1451 (Flortaucipir) is a PET ligand with high affinity for insoluble paired-helical filaments (PHFs) of hyperphosphorylated tau. In order to determine whether the development of psychosis and worsened prognosis in AD is associated with an increased burden of tau pathology that can be identified with tau imaging, we identified subjects within the Alzheimer's disease neuroimaging initiative (ADNI) who had [¹⁸F]-AV1451 imaging at baseline and became psychotic over the course of the study (N = 17) and matched them 1:3 for gender, age, and education to subjects who had [¹⁸F]-AV1451 imaging at baseline and did not become psychotic (N = 50). We compared baseline [¹⁸F]-AV1451 retention, in addition to cognitive and functional baseline and longitudinal change, in those who became psychotic over the course of participation in ADNI with those who did not. Results suggest that increases in tau pathology in frontal, medial temporal, and occipital cortices, visualized with [¹⁸F]-AV1451 binding, are associated with psychosis and a more rapid cognitive and functional decline.

Examining the role of repeated test exposure over 12 months across ADNI protocols

Objective: Changes to study protocols during longitudinal research may alter cognitive testing schedules over time. Unlike in prior Alzheimer's Disease Neuroimaging Initiative (ADNI) protocols, where testing occurred twice annually, participants enrolled in the ADNI-3 are no longer exposed to cognitive materials at 6 months. This may affect their 12-month performance relative to earlier ADNI cohorts, and potentially confounds data harmonization attempts between earlier and later ADNI protocols. Method: Using data from participants enrolled across multiple ADNI protocols, this study investigated whether test exposure during 6-month cognitive evaluation influenced scores on subsequent 12-month evaluation. Results: No interaction effects were observed between test exposure group and time at 12 months on cognitive performance. No improvements, and limited declines, were seen between baseline and 12-month follow-up scores on most measures. Conclusions: The 6-month testing session had minimal impact on 12-month performance in ADNI. Collapsing longitudinal data across ADNI protocols in future research appears appropriate.

The spike-and-slab elastic net as a classification tool in Alzheimer's disease

Alzheimer's disease (AD) is the leading cause of dementia and has received considerable research attention, including using neuroimaging biomarkers to classify patients and/or predict disease progression. Generalized linear models, e.g., logistic regression, can be used as classifiers, but since the spatial measurements are correlated and often outnumber subjects, penalized and/or Bayesian models will be identifiable, while classical models often will not. Many useful models, e.g., the elastic net and spike-and-slab lasso, perform automatic variable selection, which removes extraneous predictors and reduces model variance, but neither model exploits spatial information in selecting variables. Spatial information can be incorporated into variable selection by placing intrinsic autoregressive priors on the logit probabilities of inclusion within a spike-and-slab elastic net framework. We demonstrate the ability of this framework to improve classification performance by using cortical thickness and tau-PET images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) to classify subjects as cognitively normal or having dementia, and by using a simulation study to examine model performance using finer resolution images.

Genetic Algorithms for Optimized Diagnosis of Alzheimer's Disease and Frontotemporal Dementia Using Fluorodeoxyglucose Positron Emission Tomography Imaging

Genetic algorithms have a proven capability to explore a large space of solutions, and deal with very large numbers of input features. We hypothesized that the application of these algorithms to 18F-Fluorodeoxyglucose Positron Emission Tomography (FDG-PET) may help in diagnosis of Alzheimer's disease (AD) and Frontotemporal Dementia (FTD) by selecting the most meaningful features and automating diagnosis. We aimed to develop algorithms for the three main issues in the diagnosis: discrimination between patients with AD or FTD and healthy controls (HC), differential diagnosis between behavioral FTD (bvFTD) and AD, and differential diagnosis between primary progressive aphasia (PPA) variants. Genetic algorithms, customized with K-Nearest Neighbor and BayesNet Naives as the fitness function, were developed and compared with Principal Component Analysis (PCA). K-fold cross validation within the same sample and external validation with ADNI-3 samples were performed. External validation was performed for the algorithms distinguishing AD and HC. Our study supports the use of FDG-PET imaging, which allowed a very high accuracy rate for the diagnosis of AD, FTD, and related disorders. Genetic algorithms identified the most meaningful features with the minimum set of features, which may be relevant for automated assessment of brain FDG-PET images. Overall, our study contributes to the development of an automated, and optimized diagnosis of neurodegenerative disorders using brain metabolism.

Kappa/Lambda light-chain typing in Alzheimer's Disease

BACKGROUND

Alzheimer's disease is a progressive neurodegenerative disorder characterized by memory loss and cognitive impairment. The diagnosis of Alzheimer's disease according to symptomatic events is still a puzzling task. Developing a biomarker-based, low-cost, and high-throughput test, readily applicable in clinical laboratories, dramatically impacts the rapid and reliable detection of the disease.

OBJECTIVE

This study aimed to develop an accurate, sensitive, and reliable screening tool for diagnosing Alzheimer's disease, which can significantly reduce the cost and time of existing methods.

METHODS

We have employed a MALDI-TOF-MS-based methodology combined with a microaffinity chromatogra Results: We observed a statistically significant difference in the kappa light chain over lambda light chain (κLC/LC) ratios between patients with AD and controls (% 95 CI: -0.547 to -0.269, p<0.001). Our method demonstrated higher sensitivity (100.00%) and specificity (71.43%) for discrimination between AD and controls.

CONCLUSION

We have developed a high-throughput screening test with a novel sample enrichment method for determining κLC/LC ratios associated with AD diagnosis. Following further validation, we believe our test has a potential for clinical laboratories.

Eight-week multi-domain cognitive training does not impact large-scale resting-state brain networks in Parkinson's disease

There is meta-analytic evidence for the efficacy of cognitive training (CT) in Parkinson's disease (PD). We performed a randomized controlled trial where we found small positive effects of CT on executive function and processing speed in individuals with PD (ntotal = 140). In this study, we assessed the effects of CT on brain network connectivity and topology in a subsample of the full study population (nmri = 86). Participants were randomized into an online multi-domain CT and an active control condition and performed 24 sessions of either intervention in eight weeks. Resting-state functional MRI scans were acquired in addition to extensive clinical and neuropsychological assessments pre- and post-intervention. In line with our preregistered analysis plan (osf.io/3st82), we computed connectivity between 'cognitive' resting-state networks and computed topological outcomes at the whole-brain and sub-network level. We assessed group differences after the intervention with mixed-model analyses adjusting for baseline performance and analyzed the association between network and cognitive performance changes with repeated measures correlation analyses. The final analysis sample consisted of 71 participants (n CT = 37). After intervention there were no group differences on between-network connectivity and network topological outcomes. No associations between neural network and neuropsychological performance change were found. CT increased segregated network topology in a small sub-sample of cognitively intact participants. Post-hoc nodal analyses showed post-intervention enhanced connectivity of both the dorsal anterior cingulate cortex and dorsolateral prefrontal cortex in the CT group. The results suggest no large-scale brain network effects of eight-week computerized CT, but rather localized connectivity changes of key regions in cognitive function, that potentially reflect the specific effects of the intervention.

The Road to Personalized Medicine in Alzheimer’s Disease: The Use of Artificial Intelligence

Dementia remains an extremely prevalent syndrome among older people and represents a major cause of disability and dependency. Alzheimer’s disease (AD) accounts for the majority of dementia cases and stands as the most common neurodegenerative disease. Since age is the major risk factor for AD, the increase in lifespan not only represents a rise in the prevalence but also adds complexity to the diagnosis. Moreover, the lack of disease-modifying therapies highlights another constraint. A shift from a curative to a preventive approach is imminent and we are moving towards the application of personalized medicine where we can shape the best clinical intervention for an individual patient at a given point. This new step in medicine requires the most recent tools and analysis of enormous amounts of data where the application of artificial intelligence (AI) plays a critical role on the depiction of disease–patient dynamics, crucial in reaching early/optimal diagnosis, monitoring and intervention. Predictive models and algorithms are the key elements in this innovative field. In this review, we present an overview of relevant topics regarding the application of AI in AD, detailing the algorithms and their applications in the fields of drug discovery, and biomarkers.

Genome-Wide association study of quantitative biomarkers identifies a novel locus for alzheimer's disease at 12p12.1

BACKGROUND

Genetic study of quantitative biomarkers in Alzheimer's Disease (AD) is a promising method to identify novel genetic factors and relevant endophenotypes, which provides valuable information to deconvolute mechanistic complexity and better understand disease subtypes.

RESULTS

Using the data from the Alzheimer's Disease Neuroimaging Initiative (ADNI), we performed a genome-wide association study (GWAS) between 565,373 single nucleotide polymorphisms (SNPs) and 16 key AD biomarkers from 1,576 subjects at four visits. We identified a novel locus rs5011804 at 12p12.1 significantly associated with several AD biomarkers, including three cognitive traits (CDRSB, FAQ, ADAS13) and one imaging trait (fusiform volume). Additional mediation and interaction analyses investigated the relationships among this SNP, relevant biomarkers, and clinical diagnosis, confirming and further elaborating the genetic effects seen in the GWAS.

CONCLUSION

Our GWAS not only affirms key AD genes but also suggests the promising role of the SNP rs5011804 due to its associations with several AD cognitive and imaging outcomes. The SNP rs5011804 has a reported association with adult asthma and slightly affects intracranial volume but has not been associated with AD before. Our novel findings contribute to a more comprehensive view of the molecular mechanism behind AD.

Combined Structural MR and Diffusion Tensor Imaging Classify the Presence of Alzheimer's Disease With the Same Performance as MR Combined With Amyloid Positron Emission Tomography: A Data Integration Approach

Background: In recent years, classification frameworks using imaging data have shown that multimodal classification methods perform favorably over the use of a single imaging modality for the diagnosis of Alzheimer's Disease. The currently used clinical approach often emphasizes the use of qualitative MRI and/or PET data for clinical diagnosis. Based on the hypothesis that classification of isolated imaging modalities is not predictive of their respective value in combined approaches, we investigate whether the combination of T1 Weighted MRI and diffusion tensor imaging (DTI) can yield an equivalent performance as the combination of quantitative structural MRI (sMRI) with amyloid-PET. Methods: We parcellated the brain into regions of interest (ROI) following different anatomical labeling atlases. For each region of interest different metrics were extracted from the different imaging modalities (sMRI, PiB-PET, and DTI) to be used as features. Thereafter, the feature sets were reduced using an embedded-based feature selection method. The final reduced sets were then used as input in support vector machine (SVM) classifiers. Three different base classifiers were created, one for each imaging modality, and validated using internal (n = 41) and external data from the ADNI initiative (n = 330 for sMRI, n = 148 for DTI and n = 55 for PiB-PET) sources. Finally, the classifiers were ensembled using a weighted method in order to evaluate the performance of different combinations. Results: For the base classifiers the following performance levels were found: sMRI-based classifier (accuracy, 92%; specificity, 97% and sensitivity, 87%), PiB-PET (accuracy, 91%; specificity, 89%; and sensitivity, 92%) and the lowest performance was attained with DTI (accuracy, 80%; specificity, 76%; and sensitivity, 82%). From the multimodal approaches, when integrating two modalities, the following results were observed: sMRI+PiB-PET (accuracy, 98%; specificity, 98%; and sensitivity, 99%), sMRI+DTI (accuracy, 97%; specificity, 99%; and sensitivity, 94%) and PiB-PET+DTI (accuracy, 91%; specificity, 90%; and sensitivity, 93%). Finally, the combination of all imaging modalities yielded an accuracy of 98%, specificity of 97% and sensitivity of 99%. Conclusion: Although DTI in isolation shows relatively poor performance, when combined with structural MR, it showed a surprising classification performance which was comparable to MR combined with amyloid PET. These results are consistent with the notion that white matter changes are also important in Alzheimer's Disease.

Connectivity and variability of related cognitive subregions lead to different stages of progression toward Alzheimer's disease

Single modality MRI data is not enough to depict and discern the cause of the underlying brain pathology of Alzheimer's disease (AD). Most existing studies do not perform well with multi-group classification. To reveal the structural, functional connectivity and functional topological relationships among different stages of mild cognitive impairment (MCI) and AD, a novel method was proposed in this paper for the analysis of regional importance with an improved deep learning model. Obvious drift of related cognitive regions can be observed in the prefrontal lobe and surrounding the cingulate area in the right hemisphere when comparing AD and healthy controls (HC) based on absolute weights in the classification mode. Alterations of these regions being responsible for cognitive impairment have been previously reported. Different parcellation atlases of the human cerebral cortex were compared, and the fine-grained multimodal parcellation HCPMMP performed the best with 180 cortical areas per hemisphere. In multi-group classification, the highest accuracy achieved was 96.86% with the utilization of structural and functional topological modalities as input to the training model. Weights in the trained model with perfect discriminating ability quantify the importance of each cortical region. This is the first time such a phenomenon is discovered and weights in cortical areas are precisely described in AD and its prodromal stages to the best of our knowledge. Our findings can establish other study models to differentiate the patterns in various diseases with cognitive impairments and help to identify the underlying pathology.

Tackling a Major Deficiency of Diversity in Alzheimer's Disease Therapeutic Trials: An CTAD Task Force Report

As the last opportunity to assess treatment effect modification in a controlled setting prior to formal approval, clinical trials are a critical tool for understanding the safety and efficacy of new treatments in diverse populations. Recruitment of diverse participants in Alzheimer's Disease (AD) clinical trials are therefore essential to increase the generalizability of study results, with diversity broadly described to be representative and inclusive. This representation of study participants is equally critical in longitudinal cohort (observational) studies, which will be key to understanding disease disparities and are often used to design adequately powered AD clinical trials. New and innovative recruitment initiatives and enhanced infrastructure facilitate increased participant diversity in AD clinical studies.

Communicating and Using Dementia Risk Evidence

Advances in biomarkers, genetics, and other data used as dementia risk evidence (DRE) are increasingly informing clinical diagnosis and management. The purpose of this Mini-Forum is to provide a solutions-based discussion of the ethical and legal gaps and practical questions about how to use and communicate these data. Investigators often use DRE in research. When participants ask for their personal results, investigators have concerns. Will data that was intended to study groups be valid for individuals? Will sharing data cause distress? Debates around sharing DRE became heated when blood-based amyloid tests and amyloid reducing drugs appeared poised to enable clinicians easily to identify people with elevated brain amyloid and reduce it with a drug. Such an approach would transform the traditional role of DRE from investigational to foundational; however, then the high costs, uncertain clinical benefits and risks of the therapy led to an urgent need for education to support clinical decision making. Further complicating DRE use are direct to consumer genetic testing and increasingly available biomarker testing. Withholding DRE becomes less feasible and public education around responsible use and understanding become vital. A critical answer to these legal and ethical issues is supporting education that clearly delineates known risks, benefits, and gaps in knowledge, and communication to promote understanding among researchers, clinicians, patients, and all stakeholders. This paper provides an overview and identifies general concepts and resource documents that support more informed discussions for individuals and interdisciplinary groups.

Deep learning improves utility of tau PET in the study of Alzheimer's disease

INTRODUCTION

Positron emission tomography (PET) imaging targeting neurofibrillary tau tangles is increasingly used in the study of Alzheimer's disease (AD), but its utility may be limited by conventional quantitative or qualitative evaluation techniques in earlier disease states. Convolutional neural networks (CNNs) are effective in learning spatial patterns for image classification.

METHODS

18F-MK6240 (n = 320) and AV-1451 (n = 446) PET images were pooled from multiple studies. We performed iterations with differing permutations of radioligands, heuristics, and architectures. Performance was compared to a standard region of interest (ROI)-based approach on prediction of memory impairment. We visualized attention of the network to illustrate decision making.

RESULTS

Overall, models had high accuracy (> 80%) with good average sensitivity and specificity (75% and 82%, respectively), and had comparable or higher accuracy to the ROI standard. Visualizations of model attention highlight known characteristics of tau radioligand binding.

DISCUSSION

CNNs could improve tau PET's role in early disease and extend the utility of tau PET across generations of radioligands.

Not all voxels are created equal: Reducing estimation bias in regional NODDI metrics using tissue-weighted means

Neurite orientation dispersion and density imaging (NODDI) estimates microstructural properties of brain tissue relating to the organisation and processing capacity of neurites, which are essential elements for neuronal communication. Descriptive statistics of NODDI tissue metrics are commonly analyzed in regions-of-interest (ROI) to identify brain-phenotype associations. Here, the conventional method to calculate the ROI mean weights all voxels equally. However, this produces biased estimates in the presence of CSF partial volume. This study introduces the tissue-weighted mean, which calculates the mean NODDI metric across the tissue within an ROI, utilising the tissue fraction estimate from NODDI to reduce estimation bias. We demonstrate the proposed mean in a study of white matter abnormalities in young onset Alzheimer's disease (YOAD). Results show the conventional mean induces significant bias that correlates with CSF partial volume, primarily affecting periventricular regions and more so in YOAD subjects than in healthy controls. Due to the differential extent of bias between healthy controls and YOAD subjects, the conventional mean under- or over-estimated the effect size for group differences in many ROIs. This demonstrates the importance of using the correct estimation procedure when inferring group differences in studies where the extent of CSF partial volume differs between groups. These findings are robust across different acquisition and processing conditions. Bias persists in ROIs at higher image resolution, as demonstrated using data obtained from the third phase of the Alzheimer's disease neuroimaging initiative (ADNI); and when performing ROI analysis in template space. This suggests that conventional ROI means of NODDI metrics are biased estimates under most contemporary experimental conditions, the correction of which requires the proposed tissue-weighted mean. The tissue-weighted mean produces accurate estimates of ROI means and group differences when ROIs contain voxels with CSF partial volume. In addition to NODDI, the technique can be applied to other multi-compartment models that account for CSF partial volume, such as the free water elimination method. We expect the technique to help generate new insights into normal and abnormal variation in tissue microstructure of regions typically confounded by CSF partial volume, such as those in individuals with larger ventricles due to atrophy associated with neurodegenerative disease.

Information Flow Pattern in Early Mild Cognitive Impairment Patients

Purpose: To investigate the brain information flow pattern in patients with early mild cognitive impairment (EMCI) and explore its potential ability of differentiation and prediction for EMCI.

Methods: In this study, 49 patients with EMCI and 40 age- and sex-matched healthy controls (HCs) with available resting-state functional MRI images and neurological measures [including the neuropsychological evaluation and cerebrospinal fluid (CSF) biomarkers] were included from the Alzheimer's Disease Neuroimaging Initiative. Functional MRI measures including preferred information flow direction between brain regions and preferred information flow index of each brain region parcellated by the Atlas of Intrinsic Connectivity of Homotopic Areas (AICHA) were calculated by using non-parametric multiplicative regression-Granger causality analysis (NPMR-GCA). Edge- and node-wise Student's t-test was conducted for between-group comparison. Support vector classification was performed to differentiate EMCI from HC. The least absolute shrinkage and selection operator (lasso) regression were used to evaluate the predictive ability of information flow measures for the neurological state.

Results: Compared to HC, disturbed preferred information flow directions between brain regions involving default mode network (DMN), executive control network (ECN), somatomotor network (SMN), and visual network (VN) were observed in patients with EMCI. An altered preferred information flow index in several brain regions (including the thalamus, posterior cingulate, and precentral gyrus) was also observed. Classification accuracy of 80% for differentiating patients with EMCI from HC was achieved by using the preferred information flow directions. The preferred information flow directions have a good ability to predict memory and executive function, level of amyloid β, tau protein, and phosphorylated tau protein with the high Pearson's correlation coefficients (r > 0.7) between predictive and actual neurological measures.

Conclusion: Patients with EMCI were presented with a disturbed brain information flow pattern, which could help clinicians to identify patients with EMCI and assess their neurological state.

Large, open datasets for human connectomics research: Considerations for reproducible and responsible data use

Large, open datasets have emerged as important resources in the field of human connectomics. In this review, the evolution of data sharing involving magnetic resonance imaging is described. A summary of the challenges and progress in conducting reproducible data analyses is provided, including description of recent progress made in the development of community guidelines and recommendations, software and data management tools, and initiatives to enhance training and education. Finally, this review concludes with a discussion of ethical conduct relevant to analyses of large, open datasets and a researcher's responsibility to prevent further stigmatization of historically marginalized racial and ethnic groups. Moving forward, future work should include an enhanced emphasis on the social determinants of health, which may further contextualize findings among diverse population-based samples. Leveraging the progress to date and guided by interdisciplinary collaborations, the future of connectomics promises to be an impressive era of innovative research, yielding a more inclusive understanding of brain structure and function.

It Is Time to Study Overlapping Molecular and Circuit Pathophysiologies in Alzheimer's and Lewy Body Disease Spectra

Alzheimer's disease (AD) is the leading cause of dementia due to neurodegeneration and is characterized by extracellular senile plaques composed of amyloid β1 - 42 (Aβ) as well as intracellular neurofibrillary tangles consisting of phosphorylated tau (p-tau). Dementia with Lewy bodies constitutes a continuous spectrum with Parkinson's disease, collectively termed Lewy body disease (LBD). LBD is characterized by intracellular Lewy bodies containing α-synuclein (α-syn). The core clinical features of AD and LBD spectra are distinct, but the two spectra share common cognitive and behavioral symptoms. The accumulation of pathological proteins, which acquire pathogenicity through conformational changes, has long been investigated on a protein-by-protein basis. However, recent evidence suggests that interactions among these molecules may be critical to pathogenesis. For example, Aβ/tau promotes α-syn pathology, and α-syn modulates p-tau pathology. Furthermore, clinical evidence suggests that these interactions may explain the overlapping pathology between AD and LBD in molecular imaging and post-mortem studies. Additionally, a recent hypothesis points to a common mechanism of prion-like progression of these pathological proteins, via neural circuits, in both AD and LBD. This suggests a need for understanding connectomics and their alterations in AD and LBD from both pathological and functional perspectives. In AD, reduced connectivity in the default mode network is considered a hallmark of the disease. In LBD, previous studies have emphasized abnormalities in the basal ganglia and sensorimotor networks; however, these account for movement disorders only. Knowledge about network abnormalities common to AD and LBD is scarce because few previous neuroimaging studies investigated AD and LBD as a comprehensive cohort. In this paper, we review research on the distribution and interactions of pathological proteins in the brain in AD and LBD, after briefly summarizing their clinical and neuropsychological manifestations. We also describe the brain functional and connectivity changes following abnormal protein accumulation in AD and LBD. Finally, we argue for the necessity of neuroimaging studies that examine AD and LBD cases as a continuous spectrum especially from the proteinopathy and neurocircuitopathy viewpoints. The findings from such a unified AD and Parkinson's disease (PD) cohort study should provide a new comprehensive perspective and key data for guiding disease modification therapies targeting the pathological proteins in AD and LBD.

Frequency drift in MR spectroscopy at 3T

PURPOSE

Heating of gradient coils and passive shim components is a common cause of instability in the B0 field, especially when gradient intensive sequences are used. The aim of the study was to set a benchmark for typical drift encountered during MR spectroscopy (MRS) to assess the need for real-time field-frequency locking on MRI scanners by comparing field drift data from a large number of sites.

METHOD

A standardized protocol was developed for 80 participating sites using 99 3T MR scanners from 3 major vendors. Phantom water signals were acquired before and after an EPI sequence. The protocol consisted of: minimal preparatory imaging; a short pre-fMRI PRESS; a ten-minute fMRI acquisition; and a long post-fMRI PRESS acquisition. Both pre- and post-fMRI PRESS were non-water suppressed. Real-time frequency stabilization/adjustment was switched off when appropriate. Sixty scanners repeated the protocol for a second dataset. In addition, a three-hour post-fMRI MRS acquisition was performed at one site to observe change of gradient temperature and drift rate. Spectral analysis was performed using MATLAB. Frequency drift in pre-fMRI PRESS data were compared with the first 5:20 minutes and the full 30:00 minutes of data after fMRI. Median (interquartile range) drifts were measured and showed in violin plot. Paired t-tests were performed to compare frequency drift pre- and post-fMRI. A simulated in vivo spectrum was generated using FID-A to visualize the effect of the observed frequency drifts. The simulated spectrum was convolved with the frequency trace for the most extreme cases. Impacts of frequency drifts on NAA and GABA were also simulated as a function of linear drift. Data from the repeated protocol were compared with the corresponding first dataset using Pearson's and intraclass correlation coefficients (ICC).

RESULTS

Of the data collected from 99 scanners, 4 were excluded due to various reasons. Thus, data from 95 scanners were ultimately analyzed. For the first 5:20 min (64 transients), median (interquartile range) drift was 0.44 (1.29) Hz before fMRI and 0.83 (1.29) Hz after. This increased to 3.15 (4.02) Hz for the full 30 min (360 transients) run. Average drift rates were 0.29 Hz/min before fMRI and 0.43 Hz/min after. Paired t-tests indicated that drift increased after fMRI, as expected (p < 0.05). Simulated spectra convolved with the frequency drift showed that the intensity of the NAA singlet was reduced by up to 26%, 44 % and 18% for GE, Philips and Siemens scanners after fMRI, respectively. ICCs indicated good agreement between datasets acquired on separate days. The single site long acquisition showed drift rate was reduced to 0.03 Hz/min approximately three hours after fMRI.

DISCUSSION

This study analyzed frequency drift data from 95 3T MRI scanners. Median levels of drift were relatively low (5-min average under 1 Hz), but the most extreme cases suffered from higher levels of drift. The extent of drift varied across scanners which both linear and nonlinear drifts were observed.

Bridging the Gap Between Morphometric Similarity Mapping and Gene Transcription in Alzheimer's Disease

Disruptions in brain connectivity have been widely reported in Alzheimer’s disease (AD). Morphometric similarity (MS) mapping provides a new way of estimating structural connectivity by interregional correlation of T1WI- and DTI-derived parameters within individual brains. Here, we aimed to identify AD-related MS changing patterns and genes related to the changes and further explored the molecular and cellular mechanism underlying MS changes in AD. Both 3D-T1WI and DTI data of 106 AD patients and 106 well-matched healthy elderly individuals from the ADNI database were included in our study. Cortical regions with significantly decreased MS were found in the temporal and parietal cortex, increased MS was found in the frontal cortex and variant changes were found in the occipital cortex in AD patients. Mean MS in regions with significantly changed MS was positively or negatively associated with memory function. Negative MS-related genes were significantly downregulated in AD, specifically enriched in neurons, and participated in biological processes, with the most significant term being synaptic transmission. This study revealed AD-related cortical MS changes associated with memory function. Linking gene expression to cortical MS changes may provide a possible molecular and cellular substrate for MS abnormality and cognitive decline in AD.

Cognitive Training in Parkinson's Disease Induces Local, Not Global, Changes in White Matter Microstructure

Previous studies showed that cognitive training can improve cognitive performance in various neurodegenerative diseases but little is known about the effects of cognitive training on the brain. Here, we investigated the effects of our cognitive training paradigm, COGTIPS, on regional white matter microstructure and structural network topology. We previously showed that COGTIPS has small, positive effects on processing speed. A subsample of 79 PD patients (N = 40 cognitive training group, N = 39 active control group) underwent multi-shell diffusion-weighted imaging pre- and post-intervention. Our pre-registered analysis plan (osf.io/cht6g) entailed investigating white matter microstructural integrity (e.g., fractional anisotropy) in five tracts of interest, including the anterior thalamic radiation (ATR), whole-brain tract-based spatial statistics (TBSS), and the topology of the structural network. Relative to the active control condition, cognitive training had no effect on topology of the structural network or whole-brain TBSS. Cognitive training did lead to a reduction in fractional anisotropy in the ATR (B [SE]: - 0.32 [0.12], P = 0.01). This reduction was associated with faster responses on the Tower of London task (r = 0.42, P = 0.007), but this just fell short of our statistical threshold (P < 0.006). Post hoc "fixel-based" analyses showed that this was not due to changes in fiber density and cross section. This suggests that the observed effect in the ATR is due to training-induced alterations in neighboring fibers running through the same voxels, such as intra-striatal and thalamo-striatal fibers. These results indicate that 8 weeks of cognitive training does not alter network topology, but has subtle local effects on structural connectivity.

Data analysis with Shapley values for automatic subject selection in Alzheimer's disease data sets using interpretable machine learning

BACKGROUND

For the recruitment and monitoring of subjects for therapy studies, it is important to predict whether mild cognitive impaired (MCI) subjects will prospectively develop Alzheimer's disease (AD). Machine learning (ML) is suitable to improve early AD prediction. The etiology of AD is heterogeneous, which leads to high variability in disease patterns. Further variability originates from multicentric study designs, varying acquisition protocols, and errors in the preprocessing of magnetic resonance imaging (MRI) scans. The high variability makes the differentiation between signal and noise difficult and may lead to overfitting. This article examines whether an automatic and fair data valuation method based on Shapley values can identify the most informative subjects to improve ML classification.

METHODS

An ML workflow was developed and trained for a subset of the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort. The validation was executed for an independent ADNI test set and for the Australian Imaging, Biomarker and Lifestyle Flagship Study of Ageing (AIBL) cohort. The workflow included volumetric MRI feature extraction, feature selection, sample selection using Data Shapley, random forest (RF), and eXtreme Gradient Boosting (XGBoost) for model training as well as Kernel SHapley Additive exPlanations (SHAP) values for model interpretation.

RESULTS

The RF models, which excluded 134 of the 467 training subjects based on their RF Data Shapley values, outperformed the base models that reached a mean accuracy of 62.64% by 5.76% (3.61 percentage points) for the independent ADNI test set. The XGBoost base models reached a mean accuracy of 60.00% for the AIBL data set. The exclusion of those 133 subjects with the smallest RF Data Shapley values could improve the classification accuracy by 2.98% (1.79 percentage points). The cutoff values were calculated using an independent validation set.

CONCLUSION

The Data Shapley method was able to improve the mean accuracies for the test sets. The most informative subjects were associated with the number of ApolipoproteinE ε4 (ApoE ε4) alleles, cognitive test results, and volumetric MRI measurements.

Mitochondrial pathway polygenic risk scores are associated with Alzheimer's Disease

Genetic, animal and epidemiological studies involving biomolecular and clinical endophenotypes implicate mitochondrial dysfunction in Alzheimer's disease (AD) pathogenesis. Polygenic risk scores (PRS) provide a novel approach to assess biological pathway-associated disease risk by combining the effects of variation at multiple, functionally related genes. We investigated the associations of PRS for genes involved in 12 mitochondrial pathways (pathway-PRS) with AD in 854 participants from Alzheimer's Disease Neuroimaging Initiative. Pathway-PRS for the nuclear-encoded mitochondrial genome (OR: 1.99 [95% Cl: 1.70, 2.35]) and three mitochondrial pathways is significantly associated with increased AD risk: (i) response to oxidative stress (OR: 2.01 [95% Cl: 1.71, 2.38]); (ii) mitochondrial transport (OR: 1.81 [95% Cl: 1.55, 2.13]); (iii) hallmark oxidative phosphorylation (OR: 1.22 [95% Cl: 1.06, 1.40]. Therapeutic approaches targeting these pathways may have the potential for modifying AD pathogenesis. Further investigation is required to establish a causal role for these pathways in AD pathology.

Segmented Linear Mixed Model Analysis Reveals Association of the APOEɛ4 Allele with Faster Rate of Alzheimer's Disease Dementia Progression

Background:

APOEɛ4 allele carriers present with an increased risk for late-onset Alzheimer’s disease (AD), show cognitive symptoms at an earlier age, and are more likely to transition from mild cognitive impairment (MCI) to dementia but despite this, it remains unclear whether or not the ɛ4 allele controls the rate of disease progression.

Objective:

To determine the effects of the ɛ4 allele on rates of cognitive decline and brain atrophy during MCI and dementia stages of AD.

Methods:

A segmented linear mixed model was chosen for longitudinal modeling of cognitive and brain volumetric data of 73 ɛ3/ɛ3, 99 ɛ3/ɛ4, and 39 ɛ4/ɛ4 Alzheimer’s Disease Neuroimaging Initiative participants who transitioned during the study from MCI to AD dementia.

Results:

ɛ4 carriers showed faster decline on MMSE, ADAS-11, CDR-SB, and MoCA scales, with the last two measures showing significant ɛ4 allele-dose effects after dementia transition but not during MCI. The ɛ4 effect was more prevalent in younger participants and in females. ɛ4 carriers also demonstrated faster rates of atrophy of the whole brain, the hippocampus, the entorhinal cortex, the middle temporal gyrus, and expansion of the ventricles after transitioning to dementia but not during MCI.

Conclusion:

Possession of the ɛ4 allele is associated with a faster progression of dementia due to AD. Our observations support the notion that APOE genotype not only controls AD risk but also differentially regulates mechanisms of neurodegeneration underlying disease advancement. Furthermore, our findings carry significance for AD clinical trial design.

Towards lower-dose PET using physics-based uncertainty-aware multimodal learning with robustness to out-of-distribution data

Radiation exposure in positron emission tomography (PET) imaging limits its usage in the studies of radiation-sensitive populations, e.g., pregnant women, children, and adults that require longitudinal imaging. Reducing the PET radiotracer dose or acquisition time reduces photon counts, which can deteriorate image quality. Recent deep-neural-network (DNN) based methods for image-to-image translation enable the mapping of low-quality PET images (acquired using substantially reduced dose), coupled with the associated magnetic resonance imaging (MRI) images, to high-quality PET images. However, such DNN methods focus on applications involving test data that match the statistical characteristics of the training data very closely and give little attention to evaluating the performance of these DNNs on new out-of-distribution (OOD) acquisitions. We propose a novel DNN formulation that models the (i) underlying sinogram-based physics of the PET imaging system and (ii) the uncertainty in the DNN output through the per-voxel heteroscedasticity of the residuals between the predicted and the high-quality reference images. Our sinogram-based uncertainty-aware DNN framework, namely, suDNN, estimates a standard-dose PET image using multimodal input in the form of (i) a low-dose/low-count PET image and (ii) the corresponding multi-contrast MRI images, leading to improved robustness of suDNN to OOD acquisitions. Results on in vivo simultaneous PET-MRI, and various forms of OOD data in PET-MRI, show the benefits of suDNN over the current state of the art, quantitatively and qualitatively.

Unsupervised Machine Learning to Identify Separable Clinical Alzheimer's Disease Sub-Populations

Heterogeneity among Alzheimer's disease (AD) patients confounds clinical trial patient selection and therapeutic efficacy evaluation. This work defines separable AD clinical sub-populations using unsupervised machine learning. Clustering (t-SNE followed by k-means) of patient features and association rule mining (ARM) was performed on the ADNIMERGE dataset from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Patient sociodemographics, brain imaging, biomarkers, cognitive tests, and medication usage were included for analysis. Four AD clinical sub-populations were identified using between-cluster mean fold changes [cognitive performance, brain volume]: cluster-1 represented least severe disease [+17.3, +13.3]; cluster-0 [-4.6, +3.8] and cluster-3 [+10.8, -4.9] represented mid-severity sub-populations; cluster-2 represented most severe disease [-18.4, -8.4]. ARM assessed frequently occurring pharmacologic substances within the 4 sub-populations. No drug class was associated with the least severe AD (cluster-1), likely due to lesser antecedent disease. Anti-hyperlipidemia drugs associated with cluster-0 (mid-severity, higher volume). Interestingly, antioxidants vitamin C and E associated with cluster-3 (mid-severity, higher cognition). Anti-depressants like Zoloft associated with most severe disease (cluster-2). Vitamin D is protective for AD, but ARM identified significant underutilization across all AD sub-populations. Identification and feature characterization of four distinct AD sub-population "clusters" using standard clinical features enhances future clinical trial selection criteria and cross-study comparative analysis.

Virtual Connectomic Datasets in Alzheimer's Disease and Aging Using Whole-Brain Network Dynamics Modelling

Large neuroimaging datasets, including information about structural connectivity (SC) and functional connectivity (FC), play an increasingly important role in clinical research, where they guide the design of algorithms for automated stratification, diagnosis or prediction. A major obstacle is, however, the problem of missing features [e.g., lack of concurrent DTI SC and resting-state functional magnetic resonance imaging (rsfMRI) FC measurements for many of the subjects]. We propose here to address the missing connectivity features problem by introducing strategies based on computational whole-brain network modeling. Using two datasets, the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset and a healthy aging dataset, for proof-of-concept, we demonstrate the feasibility of virtual data completion (i.e., inferring “virtual FC” from empirical SC or “virtual SC” from empirical FC), by using self-consistent simulations of linear and nonlinear brain network models. Furthermore, by performing machine learning classification (to separate age classes or control from patient subjects), we show that algorithms trained on virtual connectomes achieve discrimination performance comparable to when trained on actual empirical data; similarly, algorithms trained on virtual connectomes can be used to successfully classify novel empirical connectomes. Completion algorithms can be combined and reiterated to generate realistic surrogate connectivity matrices in arbitrarily large number, opening the way to the generation of virtual connectomic datasets with network connectivity information comparable to the one of the original data.

Which multiband factor should you choose for your resting-state fMRI study?

Multiband acquisition, also called simultaneous multislice, has become a popular technique in resting-state functional connectivity studies. Multiband (MB) acceleration leads to a higher temporal resolution but also leads to spatially heterogeneous noise amplification, suggesting the costs may be greater in areas such as the subcortex. We evaluate MB factors of 2, 3, 4, 6, 8, 9, and 12 with 2 mm isotropic voxels, and additionally 2 mm and 3.3 mm single-band acquisitions, on a 32-channel head coil. Noise amplification was greater in deeper brain regions, including subcortical regions. Correlations were attenuated by noise amplification, which resulted in spatially varying biases that were more severe at higher MB factors. Temporal filtering decreased spatial biases in correlations due to noise amplification, but also tended to decrease effect sizes. In seed-based correlation maps, left-right putamen connectivity and thalamo-motor connectivity were highest in the single-band 3.3 mm protocol. In correlation matrices, MB 4, 6, and 8 had a greater number of significant correlations than the other acquisitions (both with and without temporal filtering). We recommend single-band 3.3 mm for seed-based subcortical analyses, and MB 4 provides a reasonable balance for studies analyzing both seed-based correlation maps and connectivity matrices. In multiband studies including secondary analyses of large-scale datasets, we recommend reporting effect sizes or test statistics instead of correlations. If correlations are reported, temporal filtering (or another method for thermal noise removal) should be used. The Emory Multiband Dataset is available on OpenNeuro.

KL-VS heterozygosity is associated with lower amyloid-dependent tau accumulation and memory impairment in Alzheimer's disease

Klotho-VS heterozygosity (KL-VS^het) is associated with reduced risk of Alzheimer’s disease (AD). However, whether KL-VS^het is associated with lower levels of pathologic tau, i.e., the key AD pathology driving neurodegeneration and cognitive decline, is unknown. Here, we assessed the interaction between KL-VS^het and levels of beta-amyloid, a key driver of tau pathology, on the levels of PET-assessed neurofibrillary tau in 551 controls and patients across the AD continuum. KL-VS^het showed lower cross-sectional and longitudinal increase in tau-PET per unit increase in amyloid-PET when compared to that of non-carriers. This association of KL-VS^het on tau-PET was stronger in Klotho mRNA-expressing brain regions mapped onto a gene expression atlas. KL-VS^het was related to better memory functions in amyloid-positive participants and this association was mediated by lower tau-PET. Amyloid-PET levels did not differ between KL-VS^het carriers versus non-carriers. Together, our findings provide evidence to suggest a protective role of KL-VS^het against amyloid-related tau pathology and tau-related memory impairments in elderly humans at risk of AD dementia.

The KL-VS haplotype of the Klotho gene has been associated with reduced risk of Alzheimer’s disease and dementia. Here the authors show an association between the KL-VS haplotype and amyloid-dependent tau accumulation using PET data.

Interplay between stress-related genes may influence Alzheimer's disease development: The results of genetic interaction analyses of human data

Emerging evidence from experimental and clinical research suggests that stress-related genes may play key roles in AD development. The fact that genome-wide association studies were not able to detect a contribution of such genes to AD indicates the possibility that these genes may influence AD non-linearly, through interactions of their products. In this paper, we selected two stress-related genes (GCN2/EIF2AK4 and APP) based on recent findings from experimental studies which suggest that the interplay between these genes might influence AD in humans. To test this hypothesis, we evaluated the effects of interactions between SNPs in these two genes on AD occurrence, using the Health and Retirement Study data on white indidividuals. We found several interacting SNP-pairs whose associations with AD remained statistically significant after correction for multiple testing. These findings emphasize the importance of nonlinear mechanisms of polygenic AD regulation that cannot be detected in traditional association studies. To estimate collective effects of multiple interacting SNP-pairs on AD, we constructed a new composite index, called Interaction Polygenic Risk Score, and showed that its association with AD is highly statistically significant. These results open a new avenue in the analyses of mechanisms of complex multigenic AD regulation.

Predicting Conversion from MCI to AD Combining Multi-Modality Data and Based on Molecular Subtype

Alzheimer’s disease (AD) is a neurodegenerative brain disease in the elderly. Identifying patients with mild cognitive impairment (MCI) who are more likely to progress to AD is a key step in AD prevention. Recent studies have shown that AD is a heterogeneous disease. In this study, we propose a subtyping-based prediction strategy to predict the conversion from MCI to AD in three years according to MCI patient subtypes. Structural magnetic resonance imaging (sMRI) data and multi-omics data, including genotype data and gene expression profiling derived from peripheral blood samples, from 125 MCI patients were used in the Alzheimer’s Disease Neuroimaging Initiative (ADNI)-1 dataset and from 98 MCI patients in the ADNI-GO/2 dataset. A variational Bayes approximation model based on the multiple kernel learning method was constructed to predict whether an MCI patient will progress to AD within three years. In internal fivefold cross-validation within ADNI-1, we achieved an overall AUC of 0.83 (79.20% accuracy, 81.25% sensitivity, 77.92% specificity) compared to the model without subtyping, which achieved an AUC of 0.78 (76.00% accuracy, 77.08% sensitivity, 75.32% specificity). In external validation using ADNI-1 as a training set and ADNI-GO/2 as an independent test set, we attained an AUC of 0.78 (74.49% accuracy, 74.19% sensitivity, 74.63% specificity). Identifying MCI patient subtypes with omics data would improve the accuracy of predicting the conversion from MCI to AD. In addition to evaluating statistics, obtaining the significant sMRI, single nucleotide polymorphism (SNP) and mRNA expression data from peripheral blood of MCI patients is noninvasive and cost-effective for predicting conversion from MCI to AD.

Genetic Influences on Hippocampal Subfields: An Emerging Area of Neuroscience Research

There is clear evidence that hippocampal subfield volumes have partly distinct genetic determinants associated with specific biological processes. The identification of genetic correlates of hippocampal subfield volumes may help to elucidate the mechanisms of neurologic diseases, as well as aging and neurodegenerative processes. However, despite the emerging interest in this area of research, the current knowledge of the genetic architecture of hippocampal subfields has not yet been consolidated. We aimed to provide a review of the current evidence from genetic studies of hippocampal subfields, highlighting current priorities and upcoming challenges. The limited number of studies investigating the influential genetic effects on hippocampal subfields, a lack of replicated results and longitudinal designs, and modest sample sizes combined with insufficient standardization of protocols are identified as the most pressing challenges in this emerging area of research.

Three-dimensional self-attention conditional GAN with spectral normalization for multimodal neuroimaging synthesis

PURPOSE

To develop a new 3D generative adversarial network that is designed and optimized for the application of multimodal 3D neuroimaging synthesis.

METHODS

We present a 3D conditional generative adversarial network (GAN) that uses spectral normalization and feature matching to stabilize the training process and ensure optimization convergence (called SC-GAN). A self-attention module was also added to model the relationships between widely separated image voxels. The performance of the network was evaluated on the data set from ADNI-3, in which the proposed network was used to predict PET images, fractional anisotropy, and mean diffusivity maps from multimodal MRI. Then, SC-GAN was applied on a multidimensional diffusion MRI experiment for superresolution application. Experiment results were evaluated by normalized RMS error, peak SNR, and structural similarity.

RESULTS

In general, SC-GAN outperformed other state-of-the-art GAN networks including 3D conditional GAN in all three tasks across all evaluation metrics. Prediction error of the SC-GAN was 18%, 24% and 29% lower compared to 2D conditional GAN for fractional anisotropy, PET and mean diffusivity tasks, respectively. The ablation experiment showed that the major contributors to the improved performance of SC-GAN are the adversarial learning and the self-attention module, followed by the spectral normalization module. In the superresolution multidimensional diffusion experiment, SC-GAN provided superior predication in comparison to 3D Unet and 3D conditional GAN.

CONCLUSION

In this work, an efficient end-to-end framework for multimodal 3D medical image synthesis (SC-GAN) is presented. The source code is also made available at https://github.com/Haoyulance/SC-GAN.

Network interdigitations of Tau and amyloid-beta deposits define cognitive levels in aging

Amyloid‐beta (Aβ) plaques and tau neurofibrillary tangles are pathological hallmarks of Alzheimer's disease (AD); their contribution to neurodegeneration and clinical manifestations are critical in understanding preclinical AD. At present, the mechanisms related to Aβ and tau pathogenesis leading to cognitive decline in older adults remain largely unknown. Here, we examined graph theory‐based positron emission tomography (PET) analytical approaches, within and between tau and Aβ PET modalities, and tested the effects on cognitive changes in cognitively normal older adults (CN). Particularly, we focused on the network interdigitations of Aβ and tau deposits, along with cognitive test scores in CN at both baseline and 2‐year follow‐up (FU). We found highly significant Aβ‐tau network integrations in AD vulnerable areas, as well as significant associations between those Aβ‐tau interdigitations and general cognitive impairment in CN at baseline and FU. Our findings suggest a distinctive contribution of interlinking network relationships between Aβ and tau deposits in heteromodal areas of the human brain. They support a network‐based interaction between Aβ and tau accumulations as a key factor for cognitive deterioration in CN prior to dementia.

Neural correlates of treatment effect and prediction of treatment outcome in patients with PTSD and comorbid personality disorder: study design

BACKGROUND

Neural alterations related to treatment outcome in patients with both post-traumatic stress disorder (PTSD) and comorbid personality disorder are unknown. Here we describe the protocol for a neuroimaging study of treatment of patients with PTSD and comorbid borderline (BPD) or cluster C (CPD) personality disorder traits. Our specific aims are to 1) investigate treatment-induced neural alterations, 2) predict treatment outcome using structural and functional magnetic resonance imaging (MRI) and 3) study neural alterations associated with BPD and CPD in PTSD patients. We hypothesize that 1) all treatment conditions are associated with normalization of limbic and prefrontal brain activity and hyperconnectivity in resting-state brain networks, with additional normalization of task-related activation in emotion regulation brain areas in the patients who receive trauma-focused therapy and personality disorder treatment; 2) Baseline task-related activation, together with structural brain measures and clinical variables predict treatment outcome; 3) dysfunction in task-related activation and resting-state connectivity of emotion regulation areas is comparable in PTSD patients with BPD or CPD, with a hypoconnected central executive network in patients with PTSD+BPD.

METHODS

We aim to include pre- and post-treatment 3 T-MRI scans in 40 patients with PTSD and (sub) clinical comorbid BPD or CPD. With an expected attrition rate of 50%, at least 80 patients will be scanned before treatment. MRI scans for 30 matched healthy controls will additionally be acquired. Patients with PTSD and BPD were randomized to either EMDR-only or EMDR combined with Dialectical Behaviour Therapy. Patients with PTSD and CPD were randomized to Imaginary Rescripting (ImRs) or to ImRs combined with Schema Focused Therapy. The scan protocol consists of a T1-weighted structural scan, resting state fMRI, task-based fMRI during an emotional face task and multi-shell diffusion weighted images. For data analysis, multivariate mixed-models, regression analyses and machine learning models will be used.

DISCUSSION

This study is one of the first to use neuroimaging measures to predict and better understand treatment response in patients with PTSD and comorbid personality disorders. A heterogeneous, naturalistic sample will be included, ensuring generalizability to a broad group of treatment seeking PTSD patients.

TRIAL REGISTRATION

Clinical Trials, NCT03833453 & NCT03833531 . Retrospectively registered, February 2019.

An MRI protocol for anatomical and functional evaluation of the California sea lion brain

BACKGROUND

Domoic acid (DOM) is a neurotoxin produced by some harmful algae blooms in coastal waters. California sea lions (Zalophus californianus) exposed to DOM often strand on beaches where they exhibit a variety of symptoms, including seizures. These animals typically show hippocampal atrophy on MRI scans.

NEW METHOD

We describe an MRI protocol for comprehensive evaluation of DOM toxicosis in the sea lion brain. We intend to study brain development in pups exposed in utero. The protocol depicts the hippocampal formation as the primary region of interest. We include scans for quantitative morphometry, functional and structural connectivity, and a cerebral blood flow map.

RESULTS

High-resolution 3D anatomical scans facilitate post hoc slicing in arbitrary planes and accurate morphometry. We demonstrate the first cerebral blood flow map using MRI, and the first structural tractography from a live sea lion brain.

COMPARISON WITH EXISTING METHODS

Scans were compared to prior anatomical and functional studies in live sea lions, and structural connectivity in post mortem specimens. Hippocampal volumes were broadly in line with prior studies, with differences likely attributable to the 3D approach used here. Functional connectivity of the dorsal left hippocampus matched that found in a prior study conducted at a lower magnetic field, while structural connectivity in the live brain agreed with findings observed in post mortem studies.

CONCLUSIONS

Our protocol provides a comprehensive, longitudinal view of the functional and anatomical changes expected to result from DOM toxicosis. It can also screen for other common neurological pathologies and is suitable for any pinniped that can fit inside an MRI scanner.

Advances in neuroimaging to support translational medicine in dementia

Advances in neuroimaging are ideally placed to facilitate the translation from progress made in cellular genetics and molecular biology of neurodegeneration into improved diagnosis, prevention and treatment of dementia. New positron emission tomography (PET) ligands allow one to quantify neuropathology, inflammation and metabolism in vivo safely and reliably, to examine mechanisms of human disease and support clinical trials. Developments in MRI-based imaging and neurophysiology provide complementary quantitative assays of brain function and connectivity, for the direct testing of hypotheses of human pathophysiology. Advances in MRI are also improving the quantitative imaging of vascular risk and comorbidities. In combination with large datasets, open data and artificial intelligence analysis methods, new informatics-based approaches are set to enable accurate single-subject inferences for diagnosis, prediction and treatment that have the potential to deliver precision medicine for dementia. Here, we show, through the use of critically appraised worked examples, how neuroimaging can bridge the gaps between molecular biology, neural circuits and the dynamics of the core systems that underpin complex behaviours. We look beyond traditional structural imaging used routinely in clinical care, to include ultrahigh field MRI (7T MRI), magnetoencephalography and PET with novel ligands. We illustrate their potential as safe, robust and sufficiently scalable to be viable for experimental medicine studies and clinical trials. They are especially informative when combined in multimodal studies, with model-based analyses to test precisely defined hypotheses.

Diagnostic accuracy of the Cogstate Brief Battery for prevalent MCI and prodromal AD (MCI A+ T+ ) in a population-based sample

INTRODUCTION

This study evaluated the diagnostic accuracy of the Cogstate Brief Battery (CBB) for mild cognitive impairment (MCI) and prodromal Alzheimer's disease (AD) in a population-based sample.

METHODS

Participants included adults ages 50+ classified as cognitively unimpaired (CU, n = 2866) or MCI (n = 226), and a subset with amyloid (A) and tau (T) positron emission tomography who were AD biomarker negative (A-T-) or had prodromal AD (A+T+).

RESULTS

Diagnostic accuracy of the Learning/Working Memory Composite (Lrn/WM) for discriminating all CU and MCI was moderate (area under the curve [AUC] = 0.75), but improved when discriminating CU A-T- and MCI A+T+ (AUC = 0.93) and when differentiating MCI participants without AD biomarkers from those with prodromal AD (AUC = 0.86). Conventional cut-offs yielded lower than expected sensitivity for both MCI (38%) and prodromal AD (73%).

DISCUSSION

Clinical utility of the CBB for detecting MCI in a population-based sample is lower than expected. Caution is needed when using currently available CBB normative data for clinical interpretation.

Mayo Normative Studies: Regression-Based Normative Data for the Auditory Verbal Learning Test for Ages 30-91 Years and the Importance of Adjusting for Sex

OBJECTIVE

Rey's Auditory Verbal Learning Test (AVLT) is a widely used word list memory test. We update normative data to include adjustment for verbal memory performance differences between men and women and illustrate the effect of this sex adjustment and the importance of excluding participants with mild cognitive impairment (MCI) from normative samples.

METHOD

This study advances the Mayo's Older Americans Normative Studies (MOANS) by using a new population-based sample through the Mayo Clinic Study of Aging, which randomly samples residents of Olmsted County, Minnesota, from age- and sex-stratified groups. Regression-based normative T-score formulas were derived from 4428 cognitively unimpaired adults aged 30-91 years. Fully adjusted T-scores correct for age, sex, and education. We also derived T-scores that correct for (1) age or (2) age and sex. Test-retest reliability data are provided.

RESULTS

From raw score analyses, sex explained a significant amount of variance in performance above and beyond age (8-10%). Applying original age-adjusted MOANS norms to the current sample resulted in significantly fewer-than-expected participants with low delayed recall performance, particularly in women. After application of new T-scores adjusted only for age, even in normative data derived from this sample, these age-adjusted T-scores showed scores <40 T occurred more frequently among men and less frequently among women relative to T-scores that also adjusted for sex.

CONCLUSIONS

Our findings highlight the importance of using normative data that adjust for sex with measures of verbal memory and provide new normative data that allow for this adjustment for the AVLT.

Detection of β-amyloid positivity in Alzheimer's Disease Neuroimaging Initiative participants with demographics, cognition, MRI and plasma biomarkers

In vivo gold standard for the ante-mortem assessment of brain β-amyloid pathology is currently β-amyloid positron emission tomography or cerebrospinal fluid measures of β-amyloid42 or the β-amyloid42/β-amyloid40 ratio. The widespread acceptance of a biomarker classification scheme for the Alzheimer's disease continuum has ignited interest in more affordable and accessible approaches to detect Alzheimer's disease β-amyloid pathology, a process that often slows down the recruitment into, and adds to the cost of, clinical trials. Recently, there has been considerable excitement concerning the value of blood biomarkers. Leveraging multidisciplinary data from cognitively unimpaired participants and participants with mild cognitive impairment recruited by the multisite biomarker study of Alzheimer's Disease Neuroimaging Initiative, here we assessed to what extent plasma β-amyloid42/β-amyloid40, neurofilament light and phosphorylated-tau at threonine-181 biomarkers detect the presence of β-amyloid pathology, and to what extent the addition of clinical information such as demographic data, APOE genotype, cognitive assessments and MRI can assist plasma biomarkers in detecting β-amyloid-positivity. Our results confirm plasma β-amyloid42/β-amyloid40 as a robust biomarker of brain β-amyloid-positivity (area under curve, 0.80-0.87). Plasma phosphorylated-tau at threonine-181 detected β-amyloid-positivity only in the cognitively impaired with a moderate area under curve of 0.67, whereas plasma neurofilament light did not detect β-amyloid-positivity in either group of participants. Clinical information as well as MRI-score independently detected positron emission tomography β-amyloid-positivity in both cognitively unimpaired and impaired (area under curve, 0.69-0.81). Clinical information, particularly APOE ε4 status, enhanced the performance of plasma biomarkers in the detection of positron emission tomography β-amyloid-positivity by 0.06-0.14 units of area under curve for cognitively unimpaired, and by 0.21-0.25 units for cognitively impaired; and further enhancement of these models with an MRI-score of β-amyloid-positivity yielded an additional improvement of 0.04-0.11 units of area under curve for cognitively unimpaired and 0.05-0.09 units for cognitively impaired. Taken together, these multi-disciplinary results suggest that when combined with clinical information, plasma phosphorylated-tau at threonine-181 and neurofilament light biomarkers, and an MRI-score could effectively identify β-amyloid+ cognitively unimpaired and impaired (area under curve, 0.80-0.90). Yet, when the MRI-score is considered in combination with clinical information, plasma phosphorylated-tau at threonine-181 and plasma neurofilament light have minimal added value for detecting β-amyloid-positivity. Our systematic comparison of β-amyloid-positivity detection models identified effective combinations of demographics, APOE, global cognition, MRI and plasma biomarkers. Promising minimally invasive and low-cost predictors such as plasma biomarkers of β-amyloid42/β-amyloid40 may be improved by age and APOE genotype.

Hippocampal subfield volumetry from structural isotropic 1 mm3 MRI scans: A note of caution

Spurred by availability of automatic segmentation software, in vivo MRI investigations of human hippocampal subfield volumes have proliferated in the recent years. However, a majority of these studies apply automatic segmentation to MRI scans with approximately 1 × 1 × 1 mm3 resolution, a resolution at which the internal structure of the hippocampus can rarely be visualized. Many of these studies have reported contradictory and often neurobiologically surprising results pertaining to the involvement of hippocampal subfields in normal brain function, aging, and disease. In this commentary, we first outline our concerns regarding the utility and validity of subfield segmentation on 1 × 1 × 1 mm3 MRI for volumetric studies, regardless of how images are segmented (i.e., manually or automatically). This image resolution is generally insufficient for visualizing the internal structure of the hippocampus, particularly the stratum radiatum lacunosum moleculare, which is crucial for valid and reliable subfield segmentation. Second, we discuss the fact that automatic methods that are employed most frequently to obtain hippocampal subfield volumes from 1 × 1 × 1 mm3 MRI have not been validated against manual segmentation on such images. For these reasons, we caution against using volumetric measurements of hippocampal subfields obtained from 1 × 1 × 1 mm3 images.

A survey on incorporating domain knowledge into deep learning for medical image analysis

Although deep learning models like CNNs have achieved great success in medical image analysis, the small size of medical datasets remains a major bottleneck in this area. To address this problem, researchers have started looking for external information beyond current available medical datasets. Traditional approaches generally leverage the information from natural images via transfer learning. More recent works utilize the domain knowledge from medical doctors, to create networks that resemble how medical doctors are trained, mimic their diagnostic patterns, or focus on the features or areas they pay particular attention to. In this survey, we summarize the current progress on integrating medical domain knowledge into deep learning models for various tasks, such as disease diagnosis, lesion, organ and abnormality detection, lesion and organ segmentation. For each task, we systematically categorize different kinds of medical domain knowledge that have been utilized and their corresponding integrating methods. We also provide current challenges and directions for future research.

Comparison of regional brain deficit patterns in common psychiatric and neurological disorders as revealed by big data

•

RVI for MDD and AD was derived based on large meta-analytical findings.

•

RVI-MDD and AD were significantly elevated in UKBB subjects with respective illnesses.

•

There was no elevation of RVI-MDD in subjects with AD or RVI-AD in subjects with MDD.

•

RVI captures neuroanatomic deviation patterns.

•

RVI is a useful biomarker for assessing similarity to neuropsychiatric illnesses.

Neurological and psychiatric illnesses are associated with regional brain deficit patterns that bear unique signatures and capture illness-specific characteristics. The Regional Vulnerability Index (RVI) was developed to quantify brain similarity by comparing individual white matter microstructure, cortical gray matter thickness and subcortical gray matter structural volume measures with neuroanatomical deficit patterns derived from large-scale meta-analytic studies. We tested the specificity of the RVI approach for major depressive disorder (MDD) and Alzheimer’s disease (AD) in a large epidemiological sample of UK Biobank (UKBB) participants (N = 19,393; 9138 M/10,255F; age = 64.8 ± 7.4 years). Compared to controls free of neuropsychiatric disorders, participants with MDD (N = 2,248; 805 M/1443F; age = 63.4 ± 7.4) had significantly higher RVI-MDD values (t = 5.6, p = 1·10⁻⁸), but showed no detectable difference in RVI-AD (t = 2.0, p = 0.10). Subjects with dementia (N = 7; 4 M/3F; age = 68.6 ± 8.6 years) showed significant elevation in RVI-AD (t = 4.2, p = 3·10⁻⁵) but not RVI-MDD (t = 2.1, p = 0.10) compared to controls. Even within affective illnesses, participants with bipolar disorder (N = 54) and anxiety disorder (N = 773) showed no significant elevation in whole-brain RVI-MDD. Participants with Parkinson’s disease (N = 37) showed elevation in RVI-AD (t = 2.4, p = 0.01) while subjects with stroke (N = 247) showed no such elevation (t = 1.1, p = 0.3). In summary, we demonstrated elevation in RVI-MDD and RVI-AD measures in the respective illnesses with strong replicability that is relatively specific to the respective diagnoses. These neuroanatomic deviation patterns offer a useful biomarker for population-wide assessments of similarity to neuropsychiatric illnesses.

A Biomarker-Centric Approach to Drug Discovery and Development: Lessons Learned from the Coronavirus Disease 2019 Pandemic

Faced with the health and economic consequences of the global spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the biomedical community came together to identify, diagnose, prevent, and treat the novel disease at breathtaking speeds. The field advanced from a publicly available viral genome to a commercialized globally scalable diagnostic biomarker test in less than 2 months, and first-in-human dosing with vaccines and repurposed antivirals followed shortly thereafter. This unprecedented efficiency was driven by three key factors: 1) international multistakeholder collaborations, 2) widespread data sharing, and 3) flexible regulatory standards tailored to meet the urgency of the situation. Learning from the remarkable success achieved during this public health crisis, we are proposing a biomarker-centric approach throughout the drug development pipeline. Although all therapeutic areas would benefit from end-to-end biomarker science, efforts should be prioritized to areas with the greatest unmet medical needs, including neurodegenerative diseases, chronic lower respiratory diseases, metabolic disorders, and malignant neoplasms. SIGNIFICANCE STATEMENT: Faced with the unprecedented threat of the severe acute respiratory syndrome coronavirus 2 pandemic, the biomedical community collaborated to develop a globally scalable diagnostic biomarker (viral DNA) that catalyzed therapeutic development at breathtaking speeds. Learning from this remarkable efficiency, we propose a multistakeholder biomarker-centric approach to drug development across therapeutic areas with unmet medical needs.

Deep learning detection of informative features in tau PET for Alzheimer's disease classification

Background

Alzheimer’s disease (AD) is the most common type of dementia, typically characterized by memory loss followed by progressive cognitive decline and functional impairment. Many clinical trials of potential therapies for AD have failed, and there is currently no approved disease-modifying treatment. Biomarkers for early detection and mechanistic understanding of disease course are critical for drug development and clinical trials. Amyloid has been the focus of most biomarker research. Here, we developed a deep learning-based framework to identify informative features for AD classification using tau positron emission tomography (PET) scans.

Results

The 3D convolutional neural network (CNN)-based classification model of AD from cognitively normal (CN) yielded an average accuracy of 90.8% based on five-fold cross-validation. The LRP model identified the brain regions in tau PET images that contributed most to the AD classification from CN. The top identified regions included the hippocampus, parahippocampus, thalamus, and fusiform. The layer-wise relevance propagation (LRP) results were consistent with those from the voxel-wise analysis in SPM12, showing significant focal AD associated regional tau deposition in the bilateral temporal lobes including the entorhinal cortex. The AD probability scores calculated by the classifier were correlated with brain tau deposition in the medial temporal lobe in MCI participants (r = 0.43 for early MCI and r = 0.49 for late MCI).

Conclusion

A deep learning framework combining 3D CNN and LRP algorithms can be used with tau PET images to identify informative features for AD classification and may have application for early detection during prodromal stages of AD.

Volumetric distribution of perivascular space in relation to mild cognitive impairment

Vascular contributions to early cognitive decline are increasingly recognized, prompting further investigation into the nature of related changes in perivascular spaces (PVS). Using magnetic resonance imaging, we show that, compared to a cognitively normal sample, individuals with early cognitive dysfunction have altered PVS presence and distribution, irrespective of Amyloid-β. Surprisingly, we noted lower PVS presence in the anterosuperior medial temporal lobe (asMTL) (1.29 times lower PVS volume fraction in cognitively impaired individuals, p < 0.0001), which was associated with entorhinal neurofibrillary tau tangle deposition (beta (standard error) = -0.98 (0.4); p = 0.014), one of the hallmarks of early Alzheimer's disease pathology. We also observed higher PVS volume fraction in centrum semi-ovale of the white matter, but only in female participants (1.47 times higher PVS volume fraction in cognitively impaired individuals, p = 0.0011). We also observed PVS changes in participants with history of hypertension (higher in the white matter and lower in the asMTL). Our results suggest that anatomically specific alteration of the PVS is an early neuroimaging feature of cognitive impairment in aging adults, which is differentially manifested in female.

Cognitive biomarker prioritization in Alzheimer's Disease using brain morphometric data

BACKGROUND

Cognitive assessments represent the most common clinical routine for the diagnosis of Alzheimer's Disease (AD). Given a large number of cognitive assessment tools and time-limited office visits, it is important to determine a proper set of cognitive tests for different subjects. Most current studies create guidelines of cognitive test selection for a targeted population, but they are not customized for each individual subject. In this manuscript, we develop a machine learning paradigm enabling personalized cognitive assessments prioritization.

METHOD

We adapt a newly developed learning-to-rank approach [Formula: see text] to implement our paradigm. This method learns the latent scoring function that pushes the most effective cognitive assessments onto the top of the prioritization list. We also extend [Formula: see text] to better separate the most effective cognitive assessments and the less effective ones.

RESULTS

Our empirical study on the ADNI data shows that the proposed paradigm outperforms the state-of-the-art baselines on identifying and prioritizing individual-specific cognitive biomarkers. We conduct experiments in cross validation and level-out validation settings. In the two settings, our paradigm significantly outperforms the best baselines with improvement as much as 22.1% and 19.7%, respectively, on prioritizing cognitive features.

CONCLUSIONS

The proposed paradigm achieves superior performance on prioritizing cognitive biomarkers. The cognitive biomarkers prioritized on top have great potentials to facilitate personalized diagnosis, disease subtyping, and ultimately precision medicine in AD.