The emergence of multiscale connectomics-based approaches in stroke recovery

Stroke is a leading cause of adult disability. Understanding stroke damage and recovery requires deciphering changes in complex brain networks across different spatiotemporal scales. While recent developments in brain readout technologies and progress in complex network modeling have revolutionized current understanding of the effects of stroke on brain networks at a macroscale, reorganization of smaller scale brain networks remains incompletely understood. In this review, we use a conceptual framework of graph theory to define brain networks from nano- to macroscales. Highlighting stroke-related brain connectivity studies at multiple scales, we argue that multiscale connectomics-based approaches may provide new routes to better evaluate brain structural and functional remapping after stroke and during recovery.

In patients with mild disability NMOSD: is the alteration in the cortical morphological or functional network topological properties more significant

Objective

To assess the alteration of individual brain morphological and functional network topological properties and their clinical significance in patients with neuromyelitis optica spectrum disorder (NMOSD).

Materials and methods

Eighteen patients with NMOSD and twenty-two healthy controls (HCs) were included. The clinical assessment of NMOSD patients involved evaluations of disability status, cognitive function, and fatigue impact. For each participant, brain images, including high-resolution T1-weighted images for individual morphological brain networks (MBNs) and resting-state functional MR images for functional brain networks (FBNs) were obtained. Topological properties were calculated and compared for both MBNs and FBNs. Then, partial correlation analysis was performed to investigate the relationships between the altered network properties and clinical variables. Finally, the altered network topological properties were used to classify NMOSD patients from HCs and to analyses time- to-progression of the patients.

Results

The average Expanded Disability Status Scale score of NMOSD patients was 1.05 (range from 0 to 2), indicating mild disability. Compared to HCs, NMOSD patients exhibited a higher normalized characteristic path length (λ) in their MBNs (P = 0.0118, FDR corrected) but showed no significant differences in the global properties of FBNs (p: 0.405-0.488). Network-based statistical analysis revealed that MBNs had more significantly altered connections (P< 0.01, NBS corrected) than FBNs. Altered nodal properties of MBNs were correlated with disease duration or fatigue scores (P< 0.05/6 with Bonferroni correction). Using the altered nodal properties of MBNs, the accuracy of classification of NMOSD patients versus HCs was 96.4%, with a sensitivity of 93.3% and a specificity of 100%. This accuracy was better than that achieved using the altered nodal properties of FBNs. Nodal properties of MBN significantly predicted Expanded Disability Status Scale worsening in patients with NMOSD.

Conclusion

The results indicated that patients with mild disability NMOSD exhibited compensatory increases in local network properties to maintain overall stability. Furthermore, the alterations in the morphological network nodal properties of NMOSD patients not only had better relevance for clinical assessments compared with functional network nodal properties, but also exhibited predictive values of EDSS worsening.

Feature attention graph neural network for estimating brain age and identifying important neural connections in mouse models of genetic risk for Alzheimer's disease

Alzheimer's disease (AD) remains one of the most extensively researched neurodegenerative disorders due to its widespread prevalence and complex risk factors. Age is a crucial risk factor for AD, which can be estimated by the disparity between physiological age and estimated brain age. To model AD risk more effectively, integrating biological, genetic, and cognitive markers is essential. Here, we utilized mouse models expressing the major APOE human alleles and human nitric oxide synthase 2 to replicate genetic risk for AD and a humanized innate immune response. We estimated brain age employing a multivariate dataset that includes brain connectomes, APOE genotype, subject traits such as age and sex, and behavioral data. Our methodology used Feature Attention Graph Neural Networks (FAGNN) for integrating different data types. Behavioral data were processed with a 2D Convolutional Neural Network (CNN), subject traits with a 1D CNN, brain connectomes through a Graph Neural Network using quadrant attention module. The model yielded a mean absolute error for age prediction of 31.85 days, with a root mean squared error of 41.84 days, outperforming other, reduced models. In addition, FAGNN identified key brain connections involved in the aging process. The highest weights were assigned to the connections between cingulum and corpus callosum, striatum, hippocampus, thalamus, hypothalamus, cerebellum, and piriform cortex. Our study demonstrates the feasibility of predicting brain age in models of aging and genetic risk for AD. To verify the validity of our findings, we compared Fractional Anisotropy (FA) along the tracts of regions with the highest connectivity, the Return-to-Origin Probability (RTOP), Return-to-Plane Probability (RTPP), and Return-to-Axis Probability (RTAP), which showed significant differences between young, middle-aged, and old age groups. Younger mice exhibited higher FA, RTOP, RTAP, and RTPP compared to older groups in the selected connections, suggesting that degradation of white matter tracts plays a critical role in aging and for FAGNN's selections. Our analysis suggests a potential neuroprotective role of APOE2, relative to APOE3 and APOE4, where APOE2 appears to mitigate age-related changes. Our findings highlighted a complex interplay of genetics and brain aging in the context of AD risk modeling.

The Effect of Physical Exercise on People with Psychosis: A Qualitative Critical Review of Neuroimaging Findings

Recently, genuine motor abnormalities have been recognized as prodromal and predictive signs of psychosis onset and progression. Therefore, physical exercise could represent a potentially relevant clinical tool in promoting the reshaping of neural connections in motor circuitry. The aim of this review is to provide an overview of the literature on neuroimaging findings as a result of physical treatment in psychosis cohorts. Twenty-one studies, all research articles, were included and discussed in this narrative review. Here, we first outlined how the psychotic brain is susceptible to structural plastic changes after aerobic physical training in pathognomic brain areas (i.e., temporal, hippocampal and parahippocampal regions). Secondly, we focused on functional changes, both region-specific and in terms of connections, to gain insights into the involvement of distant but inter-related neural regions in the plastic process occurring after treatment. Third, we attempted to bridge neural plastic changes occurring after physical interventions with clinical and cognitive outcomes of psychotic patients in order to assess the relevance of such neural reshaping in the psychiatric rehabilitation field. In conclusion, we suggest that the current state of the art is presenting physical intervention as effective in promoting neural changes for patients with psychosis; it is not only useful at the onset of the pathology but also in improving the course of the illness and its functional outcome. However, more evidence is needed to improve our knowledge of the efficacy of physical exercise in plastically reorganizing the psychotic brain in the long term, especially within regions lacking specific investigations, such as motor circuitry.

Genetic risk for schizophrenia is associated with increased proportion of indirect connections in brain networks revealed by a semi-metric analysis: evidence from population sample stratified for polygenic risk

Research studies based on tractography have revealed a prominent reduction of asymmetry in some key white-matter tracts in schizophrenia (SCZ). However, we know little about the influence of common genetic risk factors for SCZ on the efficiency of routing on structural brain networks (SBNs). Here, we use a novel recall-by-genotype approach, where we sample young adults from a population-based cohort (ALSPAC:N genotyped = 8,365) based on their burden of common SCZ risk alleles as defined by polygenic risk score (PRS). We compared 181 individuals at extremes of low (N = 91) or high (N = 90) SCZ-PRS under a robust diffusion MRI-based graph theoretical SBN framework. We applied a semi-metric analysis revealing higher SMR values for the high SCZ-PRS group compared with the low SCZ-PRS group in the left hemisphere. Furthermore, a hemispheric asymmetry index showed a higher leftward preponderance of indirect connections for the high SCZ-PRS group compared with the low SCZ-PRS group (PFDR < 0.05). These findings might indicate less efficient structural connectivity in the higher genetic risk group. This is the first study in a population-based sample that reveals differences in the efficiency of SBNs associated with common genetic risk variants for SCZ.

Intense long-term training impairs brain health compared with moderate exercise: Experimental evidence and mechanisms

The consequences of extremely intense long-term exercise for brain health remain unknown. We studied the effects of strenuous exercise on brain structure and function, its dose-response relationship, and mechanisms in a rat model of endurance training. Five-week-old male Wistar rats were assigned to moderate (MOD) or intense (INT) exercise or a sedentary (SED) group for 16 weeks. MOD rats showed the highest motivation and learning capacity in operant conditioning experiments; SED and INT presented similar results. In vivo MRI demonstrated enhanced global and regional connectivity efficiency and clustering as well as a higher cerebral blood flow (CBF) in MOD but not INT rats compared with SED. In the cortex, downregulation of oxidative phosphorylation complex IV and AMPK activation denoted mitochondrial dysfunction in INT rats. An imbalance in cortical antioxidant capacity was found between MOD and INT rats. The MOD group showed the lowest hippocampal brain-derived neurotrophic factor levels. The mRNA and protein levels of inflammatory markers were similar in all groups. In conclusion, strenuous long-term exercise yields a lesser improvement in learning ability than moderate exercise. Blunting of MOD-induced improvements in CBF and connectivity efficiency, accompanied by impaired mitochondrial energetics and, possibly, transient local oxidative stress, may underlie the findings in intensively trained rats.

Modulating brain integrative actions as a new perspective on pharmacological approaches to neuropsychiatric diseases

A critical aspect of drug development in the therapy of neuropsychiatric diseases is the "Target Problem", that is, the selection of a proper target after not simply the etiopathological classification but rather the detection of the supposed structural and/or functional alterations in the brain networks. There are novel ways of approaching the development of drugs capable of overcoming or at least reducing the deficits without triggering deleterious side effects. For this purpose, a model of brain network organization is needed, and the main aspects of its integrative actions must also be established. Thus, to this aim we here propose an updated model of the brain as a hyper-network in which i) the penta-partite synapses are suggested as key nodes of the brain hyper-network and ii) interacting cell surface receptors appear as both decoders of signals arriving to the network and targets of central nervous system diseases. The integrative actions of the brain networks follow the "Russian Doll organization" including the micro (i.e., synaptic) and nano (i.e., molecular) levels. In this scenario, integrative actions result primarily from protein-protein interactions. Importantly, the macromolecular complexes arising from these interactions often have novel structural binding sites of allosteric nature. Taking G protein-coupled receptors (GPCRs) as potential targets, GPCRs heteromers offer a way to increase the selectivity of pharmacological treatments if proper allosteric drugs are designed. This assumption is founded on the possible selectivity of allosteric interventions on G protein-coupled receptors especially when organized as "Receptor Mosaics" at penta-partite synapse level.

Distributed functional connectivity predicts neuropsychological test performance among older adults

Neuropsychological test is an essential tool in assessing cognitive and functional changes associated with late-life neurocognitive disorders. Despite the utility of the neuropsychological test, the brain-wide neural basis of the test performance remains unclear. Using the predictive modeling approach, we aimed to identify the optimal combination of functional connectivities that predicts neuropsychological test scores of novel individuals. Resting-state functional connectivity and neuropsychological tests included in the OASIS-3 dataset (n = 428) were used to train the predictive models, and the identified models were iteratively applied to the holdout internal test set (n = 216) and external test set (KSHAP, n = 151). We found that the connectivity-based predicted score tracked the actual behavioral test scores (r = 0.08-0.44). The predictive models utilizing most of the connectivity features showed better accuracy than those composed of focal connectivity features, suggesting that its neural basis is largely distributed across multiple brain systems. The discriminant and clinical validity of the predictive models were further assessed. Our results suggest that late-life neuropsychological test performance can be formally characterized with distributed connectome-based predictive models, and further translational evidence is needed when developing theoretically valid and clinically incremental predictive models.

Geodesic Distance on Optimally Regularized Functional Connectomes Uncovers Individual Fingerprints

Background: Functional connectomes (FCs) have been shown to provide a reproducible individual fingerprint, which has opened the possibility of personalized medicine for neuro/psychiatric disorders. Thus, developing accurate ways to compare FCs is essential to establish associations with behavior and/or cognition at the individual level. Methods: Canonically, FCs are compared using Pearson's correlation coefficient of the entire functional connectivity profiles. Recently, it has been proposed that the use of geodesic distance is a more accurate way of comparing FCs, one which reflects the underlying non-Euclidean geometry of the data. Computing geodesic distance requires FCs to be positive-definite and hence invertible matrices. As this requirement depends on the functional magnetic resonance imaging scanning length and the parcellation used, it is not always attainable and sometimes a regularization procedure is required. Results: In the present work, we show that regularization is not only an algebraic operation for making FCs invertible, but also that an optimal magnitude of regularization leads to systematically higher fingerprints. We also show evidence that optimal regularization is data set-dependent and varies as a function of condition, parcellation, scanning length, and the number of frames used to compute the FCs. Discussion: We demonstrate that a universally fixed regularization does not fully uncover the potential of geodesic distance on individual fingerprinting and indeed could severely diminish it. Thus, an optimal regularization must be estimated on each data set to uncover the most differentiable across-subject and reproducible within-subject geodesic distances between FCs. The resulting pairwise geodesic distances at the optimal regularization level constitute a very reliable quantification of differences between subjects.

Indirect White Matter Pathways Are Associated With Treated Naming Improvement in Aphasia

BACKGROUND

White matter disconnection of language-specific brain regions associates with worse aphasia recovery. Despite a loss of direct connections, many stroke survivors may maintain indirect connections between brain regions.

OBJECTIVE

To determine (1) whether preserved direct connections between language-specific brain regions relate to better poststroke naming treatment outcomes compared to no direct connections and (2) whether for individuals with a loss of direct connections, preserved indirect connections are associated with better treatment outcomes compared to individuals with no connections.

METHODS

We computed structural whole-brain connectomes from 69 individuals with chronic left-hemisphere stroke and aphasia who completed a 3-week-long language treatment that was supplemented by either anodal transcranial direct current stimulation (A-tDCS) or sham stimulation (S-tDCS). We determined differences in naming improvement between individuals with direct, indirect, and no connections using 1-way analyses of covariance and multivariable linear regressions.

RESULTS

Independently of tDCS modality, direct or indirect connections between the inferior frontal gyrus pars opercularis and angular gyrus were both associated with a greater increase in correct naming compared to no connections (P = .027 and P = .039, respectively). Participants with direct connections between the inferior frontal gyrus pars opercularis and middle temporal gyrus who received S-tDCS and participants with indirect connections who received A-tDCS significantly improved in naming accuracy.

CONCLUSIONS

Poststroke preservation of indirect white matter connections is associated with better treated naming improvement in aphasia even when direct connections are damaged. This mechanistic information can be used to stratify and predict treated naming recovery in individuals with aphasia.

Soft Tensor Regression

Statistical methods relating tensor predictors to scalar outcomes in a regression model generally vectorize the tensor predictor and estimate the coefficients of its entries employing some form of regularization, use summaries of the tensor covariate, or use a low dimensional approximation of the coefficient tensor. However, low rank approximations of the coefficient tensor can suffer if the true rank is not small. We propose a tensor regression framework which assumes a soft version of the parallel factors (PARAFAC) approximation. In contrast to classic PARAFAC where each entry of the coefficient tensor is the sum of products of row-specific contributions across the tensor modes, the soft tensor regression (Softer) framework allows the row-specific contributions to vary around an overall mean. We follow a Bayesian approach to inference, and show that softening the PARAFAC increases model flexibility, leads to improved estimation of coefficient tensors, more accurate identification of important predictor entries, and more precise predictions, even for a low approximation rank. From a theoretical perspective, we show that employing Softer leads to a weakly consistent posterior distribution of the coefficient tensor, irrespective of the true or approximation tensor rank, a result that is not true when employing the classic PARAFAC for tensor regression. In the context of our motivating application, we adapt Softer to symmetric and semi-symmetric tensor predictors and analyze the relationship between brain network characteristics and human traits.

Long-range fibre damage in small vessel brain disease affects aphasia severity

We sought to determine the underlying pathophysiology relating white matter hyperintensities to chronic aphasia severity. We hypothesized that: (i) white matter hyperintensities are associated with damage to fibres of any length, but to a higher percentage of long-range compared to mid- and short-range intracerebral white matter fibres; and (ii) the number of long-range fibres mediates the relationship between white matter hyperintensities and chronic post-stroke aphasia severity. We measured the severity of periventricular and deep white matter hyperintensities and calculated the number and percentages of short-, mid- and long-range white matter fibres in 48 individuals with chronic post-stroke aphasia. Correlation and mediation analyses were performed to assess the relationship between white matter hyperintensities, connectome fibre-length measures and aphasia severity as measured with the aphasia quotient of the Western Aphasia Battery-Revised (WAB-AQ). We found that more severe periventricular and deep white matter hyperintensities correlated with a lower proportion of long-range fibres (r = -0.423, P = 0.003 and r = -0.315, P = 0.029, respectively), counterbalanced by a higher proportion of short-range fibres (r = 0.427, P = 0.002 and r = 0.285, P = 0.050, respectively). More severe periventricular white matter hyperintensities also correlated with a lower proportion of mid-range fibres (r = -0.334, P = 0.020), while deep white matter hyperintensities did not correlate with mid-range fibres (r = -0.169, P = 0.250). Mediation analyses revealed: (i) a significant total effect of periventricular white matter hyperintensities on WAB-AQ (standardized beta = -0.348, P = 0.008); (ii) a non-significant direct effect of periventricular white matter hyperintensities on WAB-AQ (P > 0.05); (iii) significant indirect effects of more severe periventricular white matter hyperintensities on worse aphasia severity mediated in parallel by fewer long-range fibres (effect = -6.23, bootstrapping: standard error = 2.64, 95%CI: -11.82 to -1.56) and more short-range fibres (effect = 4.50, bootstrapping: standard error = 2.59, 95%CI: 0.16 to 10.29). We conclude that small vessel brain disease seems to affect chronic aphasia severity through a change of the proportions of long- and short-range fibres. This observation provides insight into the pathophysiology of small vessel brain disease, and its relationship with brain health and chronic aphasia severity.

Hippocampal Resting-State Functional Connectivity Patterns are More Closely Associated with Severity of Subjective Memory Decline than Whole Hippocampal and Subfield Volumes

The goal of this study was to examine whether hippocampal volume or resting-state functional connectivity (rsFC) patterns are associated with subjective memory decline (SMD) in cognitively normal aged adults. Magnetic resonance imaging data from 53 participants (mean age: 71.9 years) of the Boston University Alzheimer’s Disease Center registry were used in this cross-sectional study. Separate analyses treating SMD as a binary and continuous variable were performed. Subfield volumes were generated using FreeSurfer v6.0, and rsFC strength between the head and body of the hippocampus and the rest of the brain was calculated. Decreased left whole hippocampal volume and weaker rsFC strength between the right body of the hippocampus and the default mode network (DMN) were found in SMD+. Cognitive Change Index score was not correlated with volumetric measures but was inversely correlated with rsFC strength between the right body of the hippocampus and 6 brain networks, including the DMN, task control, and attentional networks. These findings suggest that hippocampal rsFC patterns reflect the current state of SMD in cognitively normal adults and may reflect subtle memory changes that standard neuropsychological tests are unable to capture.

In Vivo Visualization of Active Polysynaptic Circuits With Longitudinal Manganese-Enhanced MRI (MEMRI)

Manganese-enhanced magnetic resonance imaging (MEMRI) is a powerful tool for in vivo non-invasive whole-brain mapping of neuronal activity. Mn²⁺ enters active neurons via voltage-gated calcium channels and increases local contrast in T₁-weighted images. Given the property of Mn²⁺ of axonal transport, this technique can also be used for tract tracing after local administration of the contrast agent. However, MEMRI is still not widely employed in basic research due to the lack of a complete description of the Mn²⁺ dynamics in the brain. Here, we sought to investigate how the activity state of neurons modulates interneuronal Mn²⁺ transport. To this end, we injected mice with low dose MnCl₂ 2. (i.p., 20 mg/kg; repeatedly for 8 days) followed by two MEMRI scans at an interval of 1 week without further MnCl₂ injections. We assessed changes in T₁ contrast intensity before (scan 1) and after (scan 2) partial sensory deprivation (unilateral whisker trimming), while keeping the animals in a sensory enriched environment. After correcting for the general decay in Mn²⁺ content, whole brain analysis revealed a single cluster with higher signal in scan 1 compared to scan 2: the left barrel cortex corresponding to the right untrimmed whiskers. In the inverse contrast (scan 2 > scan 1), a number of brain structures, including many efferents of the left barrel cortex were observed. These results suggest that continuous neuronal activity elicited by ongoing sensory stimulation accelerates Mn²⁺ transport from the uptake site to its projection terminals, while the blockage of sensory-input and the resulting decrease in neuronal activity attenuates Mn²⁺ transport. The description of this critical property of Mn²⁺ dynamics in the brain allows a better understanding of MEMRI functional mechanisms, which will lead to more carefully designed experiments and clearer interpretation of the results.

Directed progression brain networks in Alzheimer's disease: properties and classification

This article introduces a new approach in brain connectomics aimed at characterizing the temporal spread in the brain of pathologies like Alzheimer's disease (AD). The main instrument is the development of "directed progression networks" (DPNets), wherein one constructs directed edges between nodes based on (weakly) inferred directions of the temporal spreading of the pathology. This stands in contrast to many previously studied brain networks where edges represent correlations, physical connections, or functional progressions. In addition, this is one of a few studies showing the value of using directed networks in the study of AD. This article focuses on the construction of DPNets for AD using longitudinal cortical thickness measurements from magnetic resonance imaging data. The network properties are then characterized, providing new insights into AD progression, as well as novel markers for differentiating normal cognition (NC) and AD at the group level. It also demonstrates the important role of nodal variations for network classification (i.e., the significance of standard deviations, not just mean values of nodal properties). Finally, the DPNets are utilized to classify subjects based on their global network measures using a variety of data-mining methodologies. In contrast to most brain networks, these DPNets do not show high clustering and small-world properties.