Profiling heterogeneity of Alzheimer's disease using white-matter impairment factors

A review of neuroimaging-based data-driven approach for Alzheimer's disease heterogeneity analysis

Alzheimer's disease (AD) is a complex form of dementia and due to its high phenotypic variability, its diagnosis and monitoring can be quite challenging. Biomarkers play a crucial role in AD diagnosis and monitoring, but interpreting these biomarkers can be problematic due to their spatial and temporal heterogeneity. Therefore, researchers are increasingly turning to imaging-based biomarkers that employ data-driven computational approaches to examine the heterogeneity of AD. In this comprehensive review article, we aim to provide health professionals with a comprehensive view of past applications of data-driven computational approaches in studying AD heterogeneity and planning future research directions. We first define and offer basic insights into different categories of heterogeneity analysis, including spatial heterogeneity, temporal heterogeneity, and spatial-temporal heterogeneity. Then, we scrutinize 22 articles relating to spatial heterogeneity, 14 articles relating to temporal heterogeneity, and five articles relating to spatial-temporal heterogeneity, highlighting the strengths and limitations of these strategies. Furthermore, we discuss the importance of understanding spatial heterogeneity in AD subtypes and their clinical manifestations, biomarkers for abnormal orderings and AD stages, the recent advancements in spatial-temporal heterogeneity analysis for AD, and the emerging role of omics data integration in advancing personalized diagnosis and treatment for AD patients. By emphasizing the significance of understanding AD heterogeneity, we hope to stimulate further research in this field to facilitate the development of personalized interventions for AD patients.

Sparse Deep Neural Network for Encoding and Decoding the Structural Connectome

Brain state classification by applying deep learning techniques on neuroimaging data has become a recent topic of research. However, unlike domains where the data is low dimensional or there are large number of available training samples, neuroimaging data is high dimensional and has few training samples. To tackle these issues, we present a sparse feedforward deep neural architecture for encoding and decoding the structural connectome of the human brain. We use a sparsely connected element-wise multiplication as the first hidden layer and a fixed transform layer as the output layer. The number of trainable parameters and the training time is significantly reduced compared to feedforward networks. We demonstrate superior performance of this architecture in encoding the structural connectome implicated in Alzheimer's disease (AD) and Parkinson's disease (PD) from DTI brain scans. For decoding, we propose recursive feature elimination (RFE) algorithm based on DeepLIFT, layer-wise relevance propagation (LRP), and Integrated Gradients (IG) algorithms to remove irrelevant features and thereby identify key biomarkers associated with AD and PD. We show that the proposed architecture reduces 45.1% and 47.1% of the trainable parameters compared to a feedforward DNN with an increase in accuracy by 2.6 % and 3.1% for cognitively normal (CN) vs AD and CN vs PD classification, respectively. We also show that the proposed RFE method leads to a further increase in accuracy by 2.1% and 4% for CN vs AD and CN vs PD classification, while removing approximately 90% to 95% irrelevant features. Furthermore, we argue that the biomarkers (i.e., key brain regions and connections) identified are consistent with previous literature. We show that relevancy score-based methods can yield high discriminative power and are suitable for brain decoding. We also show that the proposed approach led to a reduction in the number of trainable network parameters, an increase in classification accuracy, and a detection of brain connections and regions that were consistent with earlier studies.

An eXplainability Artificial Intelligence approach to brain connectivity in Alzheimer's disease

The advent of eXplainable Artificial Intelligence (XAI) has revolutionized the way human experts, especially from non-computational domains, approach artificial intelligence; this is particularly true for clinical applications where the transparency of the results is often compromised by the algorithmic complexity. Here, we investigate how Alzheimer's disease (AD) affects brain connectivity within a cohort of 432 subjects whose T1 brain Magnetic Resonance Imaging data (MRI) were acquired within the Alzheimer's Disease Neuroimaging Initiative (ADNI). In particular, the cohort included 92 patients with AD, 126 normal controls (NC) and 214 subjects with mild cognitive impairment (MCI). We show how graph theory-based models can accurately distinguish these clinical conditions and how Shapley values, borrowed from game theory, can be adopted to make these models intelligible and easy to interpret. Explainability analyses outline the role played by regions like putamen, middle and superior temporal gyrus; from a class-related perspective, it is possible to outline specific regions, such as hippocampus and amygdala for AD and posterior cingulate and precuneus for MCI. The approach is general and could be adopted to outline how brain connectivity affects specific brain regions.

Altered structural and functional homotopic connectivity associated with the progression from mild cognitive impairment to Alzheimer's disease

The progressive mild cognitive impairment (pMCI) is associated with an increased risk of Alzheimer's disease (AD). Many studies have reported the disrupted brain alteration during the imminent conversion from pMCI to AD. However, the subtle difference of structural and functional of inter-hemispheric between pMCI and stable mild cognitive impairment (sMCI) remains unknown. In the present study, we scanned the multimodal magnetic resonance imaging of 38 sMCI, 26 pMCI, and 50 healthy controls (HC) and assessed the cognitive function. The voxel-mirrored homotopic connectivity (VMHC) and volume of corpus callosum were calculated. A structural equation modeling (SEM) was established to determine the relationships between the corpus callosum, the inter-hemispheric connectivity, and cognitive assessment. Compared to sMCI, pMCI exhibited decreased VMHC in insular and thalamus, and reduced volume of corpus callosum. SEM results showed that decreased inter-hemispheric connectivity was directly associated with cognitive impairment and corpus callosum atrophy, and corpus callosum atrophy indirectly caused cognitive impairment by mediating inter-hemispheric connectivity in pMCI. In conclusion, the destruction of homotopic connectivity is related to cognitive impairment, and the corpus callosum atrophy partially mediates the association between the homotopic connectivity and cognitive impairment in pMCI.

Fecal Bile Acids and Neutral Sterols Are Associated with Latent Microbial Subgroups in the Human Gut

Bile acids, neutral sterols, and the gut microbiome are intricately intertwined and each affects human health and metabolism. However, much is still unknown about this relationship. This analysis included 1280 participants of the KORA FF4 study. Fecal metabolites (primary and secondary bile acids, plant and animal sterols) were analyzed using a metabolomics approach. Dirichlet regression models were used to evaluate associations between the metabolites and twenty microbial subgroups that were previously identified using latent Dirichlet allocation. Significant associations were identified between 12 of 17 primary and secondary bile acids and several of the microbial subgroups. Three subgroups showed largely positive significant associations with bile acids, and six subgroups showed mostly inverse associations with fecal bile acids. We identified a trend where microbial subgroups that were previously associated with “healthy” factors were here inversely associated with fecal bile acid levels. Conversely, subgroups that were previously associated with “unhealthy” factors were positively associated with fecal bile acid levels. These results indicate that further research is necessary regarding bile acids and microbiota composition, particularly in relation to metabolic health.

Four Distinct Subtypes of Alzheimer's Disease Based on Resting-State Connectivity Biomarkers

BACKGROUND

Alzheimer's disease (AD) is a neurodegenerative disorder with significant heterogeneity. Different AD phenotypes may be associated with specific brain network changes. Uncovering disease heterogeneity by using functional networks could provide insights into precise diagnoses.

METHODS

We investigated the subtypes of AD using nonnegative matrix factorization clustering on the previously identified 216 resting-state functional connectivities that differed between AD and normal control subjects. We conducted the analysis using a discovery dataset (n = 809) and a validated dataset (n = 291). Next, we grouped individuals with mild cognitive impairment according to the model obtained in the AD groups. Finally, the clinical measures and brain structural characteristics were compared among the subtypes to assess their relationship with differences in the functional network.

RESULTS

Individuals with AD were clustered into 4 subtypes reproducibly, which included those with 1) diffuse and mild functional connectivity disruption (subtype 1), 2) predominantly decreased connectivity in the default mode network accompanied by an increase in the prefrontal circuit (subtype 2), 3) predominantly decreased connectivity in the anterior cingulate cortex accompanied by an increase in prefrontal cortex connectivity (subtype 3), and 4) predominantly decreased connectivity in the basal ganglia accompanied by an increase in prefrontal cortex connectivity (subtype 4). In addition to these differences in functional connectivity, differences between the AD subtypes were found in cognition, structural measures, and cognitive decline patterns.

CONCLUSIONS

These comprehensive results offer new insights that may advance precision medicine for AD and facilitate strategies for future clinical trials.

Pathobiological Subtypes of Alzheimer Disease

Alzheimer disease (AD), the most common form of dementia, is a heterogenous disorder with various pathobiological subtypes. In addition to the 4 major subtypes based on the distribution of tau pathology and brain atrophy (typical, limbic predominant, hippocampal sparing, and minimal atrophy [MA]), several other clinical variants showing distinct regional patterns of tau burden have been identified: nonamnestic, corticobasal syndromal, primary progressive aphasia, posterior cortical atrophy, behavioral/dysexecutive, and mild dementia variants. Among the subtypes, differences were found in age at onset, sex distribution, cognitive status, disease duration, APOE genotype, and biomarker levels. The patterns of key network destructions parallel the tau and atrophy patterns of the AD subgroups essentially. Interruption of key networks, in particular the default-mode network that is responsible for cognitive decline, is consistent in hetero-genous AD groups. AD pathology is often associated with co-pathologies: cerebrovascular lesions, Lewy pathology, and TDP-43 proteinopathies. These mixed pathologies essentially influence the clinical picture of AD and may accel-erate disease progression. Unraveling the heterogeneity among the AD spectrum entities is important for opening a window to pathogenic mechanisms affecting the brain and enabling precision medicine approaches as a basis for developing preventive and ultimately successful disease-modifying therapies for AD.

A Review of Brain Atrophy Subtypes Definition and Analysis for Alzheimer’s Disease Heterogeneity Studies

Alzheimer’s disease (AD) is a heterogeneous disease with different subtypes. Studying AD subtypes from brain structure, neuropathology, and cognition are of great importance for AD heterogeneity research. Starting from the study of constructing AD subtypes based on the features of T1-weighted structural magnetic resonance imaging, this paper introduces the major connections between the subtype definition and analysis strategies, including brain region-based subtype definition, and their demographic, neuropathological, and neuropsychological characteristics. The advantages and existing problems are analyzed, and reasonable improvement schemes are prospected. Overall, this review offers a more comprehensive view in the field of atrophy subtype in AD, along with their advantages, challenges, and future prospects, and provide a basis for improving individualized AD diagnosis.

Associations between habitual diet, metabolic disease, and the gut microbiota using latent Dirichlet allocation

BACKGROUND

The gut microbiome impacts human health through various mechanisms and is involved in the development of a range of non-communicable diseases. Diet is a well-known factor influencing microbe-host interaction in health and disease. However, very few findings are based on large-scale analysis using population-based studies. Our aim was to investigate the cross-sectional relationship between habitual dietary intake and gut microbiota structure in the Cooperative Health Research in the Region of Augsburg (KORA) FF4 study.

RESULTS

Fecal microbiota was analyzed using 16S rRNA gene amplicon sequencing. Latent Dirichlet allocation (LDA) was applied to samples from 1992 participants to identify 20 microbial subgroups within the study population. Each participant's gut microbiota was subsequently described by a unique composition of these 20 subgroups. Associations between habitual dietary intake, assessed via repeated 24-h food lists and a Food Frequency Questionnaire, and the 20 subgroups, as well as between prevalence of metabolic diseases/risk factors and the subgroups, were assessed with multivariate-adjusted Dirichlet regression models. After adjustment for multiple testing, eight of 20 microbial subgroups were significantly associated with habitual diet, while nine of 20 microbial subgroups were associated with the prevalence of one or more metabolic diseases/risk factors. Subgroups 5 (Faecalibacterium, Lachnospiracea incertae sedis, Gemmiger, Roseburia) and 14 (Coprococcus, Bacteroides, Faecalibacterium, Ruminococcus) were particularly strongly associated with diet. For example, participants with a high probability for subgroup 5 were characterized by a higher Alternate Healthy Eating Index and Mediterranean Diet Score and a higher intake of food items such as fruits, vegetables, legumes, and whole grains, while participants with prevalent type 2 diabetes mellitus were characterized by a lower probability for subgroup 5.

CONCLUSIONS

The associations between habitual diet, metabolic diseases, and microbial subgroups identified in this analysis not only expand upon current knowledge of diet-microbiota-disease relationships, but also indicate the possibility of certain microbial groups to be modulated by dietary intervention, with the potential of impacting human health. Additionally, LDA appears to be a powerful tool for interpreting latent structures of the human gut microbiota. However, the subgroups and associations observed in this analysis need to be replicated in further studies. Video abstract.

Fronto-parietal networks underlie the interaction between executive control and conscious perception: Evidence from TMS and DWI

The executive control network is involved in novel situations or those in which prepotent responses need to be overridden. Previous studies have demonstrated that when control is exerted, conscious perception is impaired, and this effect is related to the functional connectivity of fronto-parietal regions. In the present study, we explored the causal involvement of one of the nodes of this fronto-parietal network (the right Supplementary Motor Area, SMA) in the interaction between executive control and conscious perception. Participants performed a dual task in which they responded to a Stroop task while detecting the presence/absence of a near-threshold Gabor stimulus. Concurrently, transcranial magnetic stimulation (TMS) was applied over the right SMA or a control site (vertex; Experiment 1). As a further control, the right Frontal Eye Field (FEF) was stimulated in Experiment 2. Diffusion-weighted imaging (DWI) tractography was used to isolate the three branches of the superior longitudinal fasciculus (SLF I, II and III), and the frontal aslant tract (FAT), and to explore if TMS effects were related to their micro- and macrostructural characteristics. Results demonstrated reduced perceptual sensitivity on incongruent as compared to congruent Stroop trials. A causal role of the right SMA on the modulation of perceptual sensitivity by executive control was only demonstrated when the microstructure of the right SLF III or the left FAT were taken into account. The volume of the right SLF III was also related to the modulation of response criterion by executive control when the right FEF was stimulated. These results add evidence in favor of shared neural correlates for attention and conscious perception in fronto-parietal regions and highlight the role of white matter in TMS effects.

Disentangling Heterogeneity in Alzheimer's Disease and Related Dementias Using Data-Driven Methods

Brain aging is a complex process that includes atrophy, vascular injury, and a variety of age-associated neurodegenerative pathologies, together determining an individual's course of cognitive decline. While Alzheimer's disease and related dementias contribute to the heterogeneity of brain aging, these conditions themselves are also heterogeneous in their clinical presentation, progression, and pattern of neural injury. We reviewed studies that leveraged data-driven approaches to examining heterogeneity in Alzheimer's disease and related dementias, with a principal focus on neuroimaging studies exploring subtypes of regional neurodegeneration patterns. Over the past decade, the steadily increasing wealth of clinical, neuroimaging, and molecular biomarker information collected within large-scale observational cohort studies has allowed for a richer understanding of the variability of disease expression within the aging and Alzheimer's disease and related dementias continuum. Moreover, the availability of these large-scale datasets has supported the development and increasing application of clustering techniques for studying disease heterogeneity in a data-driven manner. In particular, data-driven studies have led to new discoveries of previously unappreciated disease subtypes characterized by distinct neuroimaging patterns of regional neurodegeneration, which are paralleled by heterogeneous profiles of pathological, clinical, and molecular biomarker characteristics. Incorporating these findings into novel frameworks for more differentiated disease stratification holds great promise for improving individualized diagnosis and prognosis of expected clinical progression, and provides opportunities for development of precision medicine approaches for therapeutic intervention. We conclude with an account of the principal challenges associated with data-driven heterogeneity analyses and outline avenues for future developments in the field.

Biological subtypes of Alzheimer disease: A systematic review and meta-analysis

OBJECTIVE

To test the hypothesis that distinct subtypes of Alzheimer disease (AD) exist and underlie the heterogeneity within AD, we conducted a systematic review and meta-analysis on AD subtype studies based on postmortem and neuroimaging data.

METHODS

EMBASE, PubMed, and Web of Science databases were consulted until July 2019.

RESULTS

Neuropathology and neuroimaging studies have consistently identified 3 subtypes of AD based on the distribution of tau-related pathology and regional brain atrophy: typical, limbic-predominant, and hippocampal-sparing AD. A fourth subtype, minimal atrophy AD, has been identified in several neuroimaging studies. Typical AD displays tau-related pathology and atrophy both in hippocampus and association cortex and has a pooled frequency of 55%. Limbic-predominant, hippocampal-sparing, and minimal atrophy AD had a pooled frequency of 21%, 17%, and 15%, respectively. Between-subtype differences were found in age at onset, age at assessment, sex distribution, years of education, global cognitive status, disease duration, APOE ε4 genotype, and CSF biomarker levels.

CONCLUSION

We identified 2 core dimensions of heterogeneity: typicality and severity. We propose that these 2 dimensions determine individuals' belonging to one of the AD subtypes based on the combination of protective factors, risk factors, and concomitant non-AD brain pathologies. This model is envisioned to aid with framing hypotheses, study design, interpretation of results, and understanding mechanisms in future subtype studies. Our model can be used along the A/T/N classification scheme for AD biomarkers. Unraveling the heterogeneity within AD is critical for implementing precision medicine approaches and for ultimately developing successful disease-modifying drugs for AD.

White Matter Microstructural Damage as an Early Sign of Subjective Cognitive Decline

Background and Objective: Subjective cognitive decline (SCD) is considered a preclinical state of Alzheimer's disease (AD) and may represent a more advanced preclinical status than amnestic mild cognitive impairment (aMCI). Our aim was to explore changes in the white matter (WM) microstructure and their correlation with cognitive function in these AD-spectrum patients. Methods: Diffusion tensor images from 43 individuals with normal cognition (NC), 38 SCD patients, and 36 aMCI patients were compared using an atlas-based segmentation strategy. The correlation between diffusion parameters and cognitive function was further analyzed. Results: The anatomical pattern of WM impairment was generally similar between SCD and aMCI patients. However, aMCI patients showed significantly lower fractional anisotropy (i.e., corpus callosum forceps major and forceps minor) and increased mean diffusivity [i.e., bilateral anterior thalamic radiation (ATR), left corticospinal tract (CST), forceps minor, left cingulum (cingulate gyrus), left cingulum hippocampus, and left inferior fronto-occipital fasciculus (IFO)] in some tracts than did SCD subjects, indicating a disruption in WM microstructural integrity in the aMCI. Individuals with microstructural disruption in forceps minor, left cingulum (cingulate gyrus), and left cingulum hippocampus tracts performed worse in general cognition and memory function tests, as indicated by line regression analysis. Conclusion: SCD individuals had extensive WM microstructural damage in a pattern similar to that seen in aMCI, although presenting a cognitive performance comparable with that of cognitively healthy individuals. Our results suggest that WM integrity might precede objectively measurable memory decline and may be a potential early biomarker for AD.