Amyloid- β (A β ) plaques in conjunction with hyperphosphorylated tau proteins in the form of neurofibrillary tangles are the two neuropathological hallmarks of Alzheimer’s disease. It is well-known that the identification of individuals with A β positivity could enable early diagnosis. In this work, we aim at capturing the A β positivity status in an unbalanced cohort enclosing subjects at different disease stages, exploiting the underlying structural and connectivity disease-induced modulations as revealed by structural, functional, and diffusion MRI. Of note, due to the unbalanced cohort, the outcomes may be guided by those factors rather than amyloid accumulation. The partial views provided by each modality are integrated in the model, allowing to take full advantage of their complementarity in encoding the effects of the A β accumulation, leading to an accuracy of 0.762 ± 0.04. The specificity of the information brought by each modality is assessed by post hoc explainability analysis (guided backpropagation), highlighting the underlying structural and functional changes. Noteworthy, well-established biomarker key regions related to A β deposition could be identified by all modalities, including the hippocampus, thalamus, precuneus, and cingulate gyrus, witnessing in favor of the reliability of the method as well as its potential in shedding light on modality-specific possibly unknown A β deposition signatures. Author Summary In this work, we employed a multimodal MRI-based deep learning framework for the classification of unbalanced cohorts relying on the amyloid- β status in the Alzheimer’s disease continuum. To this end, structural, functional, and diffusion MRI data were used to feed a 3D-CNN and two different graph neural networks, respectively, reaching an accuracy of 0.762 ± 0.04. Post hoc explainability analysis was performed to extract the most relevant regions that led to the outcome, highlighting the involvement of different cortical and subcortical regions. This work provides evidence of the added value brought by exploiting different imaging modalities in decrypting the nature and extent of brain alterations in the amyloid-guided classification outcome.

Promising evidence has suggested potential links between mind-wandering and Alzheimer’s disease (AD). Yet, older adults with diagnosable neurocognitive disorders show reduced meta-awareness, thus questioning the validity of probe-assessed mind-wandering in older adults. In prior work, we employed response time variability as an objective, albeit indirect, marker of mind-wandering to identify patterns of functional connectivity that predicted mind-wandering. In the current study, we evaluated the association of this connectome-based, mind-wandering model with cerebral spinal fluid (CSF) p-tau/A β 42 ratio in 289 older adults from the Alzheimer’s Disease NeuroImaging Initiative (ADNI). Moreover, we examined if this model was similarly associated with individual differences in composite measures of global cognition, episodic memory, and executive functioning. Edges from the high response time variability model were significantly associated with CSF p-tau/A β ratio. Furthermore, connectivity strength within edges associated with high response time variability was negatively associated with global cognition and episodic memory functioning. This study provides the first empirical support for a link between an objective neuromarker of mind-wandering and AD pathophysiology. Given the observed association between mind-wandering and cognitive functioning in older adults, interventions targeted at reducing mind-wandering, particularly before the onset of AD pathogenesis, may make a significant contribution to the prevention of AD-related cognitive decline. Author Summary Response time variability is considered an objective, albeit indirect, marker of mind-wandering. In this study, we applied a previously-derived connectome-based model of response time variability to resting-state data obtained from 289 older adults in the Alzheimer’s Disease NeuroImaging Initiative. The network strength of the high response time variability model was correlated with a cerebrospinal fluid (CSF)-based ratiometric measure of amyloid and tau pathology. Additionally, our results demonstrated that the network strength in the high response time variability model was also linked with global cognition and episodic memory. This study provides the first empirical support for the association between a neuromarker of response time variability—an indirect marker of mind-wandering—and AD pathophysiology.

Amyloid-beta (Aβ) deposition and altered brain structure are the most relevant neuroimaging biomarkers for Alzheimer’s disease (AD). However, their spatial inconsistency was always confusing and misleading. Furthermore, the relationship between this spatial inconsistency and AD progression is unclear. The current study introduced a regional radiomics similarity network (R2SN) to map structural MRI and Aβ positron emission tomography (PET) images to study their cross-modal interregional coupling. A total of 790 participants (248 normal controls, 390 mild cognitive impaired patients, and 152 AD patients) with their structural MRI and PET images were studied. The results showed that global and regional R2SN coupling significantly decreased according to the severity of cognitive decline, from mild cognitive impairment to AD dementia. The global coupling patterns are discriminative between different APOE ε4, Aβ, and Tau subgroups. R2SN coupling was probed for relationships with neuropsychiatric measures and peripheral biomarkers. Kaplan–Meier analysis showed that lower global coupling scores could reveal worse clinical progression of dementia. The R2SN coupling scores derived from the coupling between Aβ and atrophy over individual brain regions could reflect the specific pathway of AD progression, which would be a reliable biomarker for AD. Author Summary Amyloid-beta (Aβ) deposition and altered brain structure are the most relevant neuroimaging biomarkers for Alzheimer’s disease (AD). We introduced a novel network coupling measure based on the regional radiomics similarity network (R2SN) to explore the potential association between the spatial distributions of brain structure and Aβ based on sMRI and Aβ positron emission tomography (PET) imaging. In this study, we systematically demonstrated that the alteration of the coupling between brain networks of brain structure and Aβ accumulation could serve as a predictor for revealing the distinct progression of AD.

A hallmark of Alzheimer’s disease is the aggregation of insoluble amyloid-beta plaques and tau protein neurofibrillary tangles. A key histopathological observation is that tau protein aggregates follow a structured progression pattern through the brain. Mathematical network models of prion-like propagation have the ability to capture such patterns, but a number of factors impact the observed staging result, thus introducing questions regarding model selection. Here, we introduce a novel approach, based on braid diagrams, for studying the structured progression of a marker evolving on a network. We apply this approach to a six-stage ‘Braak pattern’ of tau proteins, in Alzheimer’s disease, motivated by a recent observation that seed-competent tau precedes tau aggregation. We show that the different modeling choices, from the model parameters to the connectome resolution, play a significant role in the landscape of observable staging patterns. Our approach provides a systematic way to approach model selection for network propagation of neurodegenerative diseases that ensures both reproducibility and optimal parameter fitting. Author Summary Network diffusion models of neurodegenerative diseases are a class of dynamical systems that simulate the evolution of toxic proteins on the connectome. These models predict, from an initial seed, a pattern of invasion called staging. The generalized staging problem seeks to systematically study the effect of various model choices on staging. We introduce methods based on braid diagrams to test the possible staging landscape of a model and how it depends on the choice of connectome, as well as the model parameters. Our primary finding is that connectome construction, the choice of the graph Laplacian, and transport models all have an impact on staging that should be taken into account in any study.