Resting-state Brain Connectivity Research Articles

Physiotherapy-assisted overground exoskeleton use: mixed methods feasibility study protocol quantifying the user experience, as well as functional, neural, and muscular outcomes in children with mobility impairments.

Early phase research suggests that physiotherapy paired with use of robotic walking aids provides a novel opportunity for children with severe mobility challenges to experience active walking. The Trexo Plus is a pediatric lower limb exoskeleton mounted on a wheeled walker frame, and is adjustable to fit a child's positional and gait requirements. It guides and powers the child's leg movements in a way that is individualized to their movement potential and upright support needs, and can provide progressive challenges for walking within a physiotherapy-based motor learning treatment paradigm. This protocol outlines a single group mixed-methods study that assesses the feasibility of physiotherapy-assisted overground Trexo use in school and outpatient settings during a 6-week physiotherapy block. Children ages 3-6 years (n = 10; cerebral palsy or related disorder, Gross Motor Function Classification System level IV) will be recruited by circle of care invitations to participate. Study indicators/outcomes will focus on evaluation of: (i) clinical feasibility, safety, and acceptability of intervention; (ii) pre-post intervention motor/functional outcomes; (iii) pre-post intervention brain structure characterization and resting state brain connectivity; (iv) muscle activity characterization during Trexo-assisted gait and natural assisted gait; (v) heart rate during Trexo-assisted gait and natural assisted gait; and (vi) user experience and perceptions of physiotherapists, children, and parents. This will be the first study to investigate feasibility indicators, outcomes, and experiences of Trexo-based physiotherapy in a school and outpatient context with children who have mobility challenges. It will explore the possibility of experience-dependent neuroplasticity in the context of gait rehabilitation, as well as associated functional and muscular outcomes. Finally, the study will address important questions about clinical utility and future adoption of the device from the physiotherapists' perspective, comfort and engagement from the children's perspective, and the impressions of parents about the value of introducing this technology as an early intervention. https://clinicaltrials.gov, identifier NCT05463211.

Identifying the Shared and Dissociable Neural Bases between Self-Worth and Moral Ambivalence.

Self-ambivalence, a prevalent phenomenon in daily life, has been increasingly substantiated by research. It refers to conflicting self-views and evaluations, primarily concerning self-worth and morality. Previous behavioral research has distinguished self-worth and moral ambivalence, but it remains unclear whether they have separable neural bases. The present study addressed this question by examining resting-state brain activity (i.e., the fractional amplitude of low-frequency fluctuations, fALFF) and connectivity (i.e., resting-state functional connectivity, RSFC) in 112 college students. The results found that self-worth ambivalence was positively related to the fALFF in the orbitofrontal cortex (OFC) and left superior parietal lobule (SPL). The RSFC strength between the SPL and precuneus/posterior cingulate cortex (PCC) was positively related to self-worth ambivalence. Moral ambivalence was positively associated with the fALFF in the left SPL (extending into the temporoparietal junction) and right SPL. The RSFC strengths between the left SPL/TPJ and OFC, as well as the RSFC strengths between the right SPL as a seed and the bilateral middle and inferior temporal gyrus, were associated with moral ambivalence. Overall, the neural bases of self-worth and moral ambivalence are associated with the SPL and OFC, involved in attentional alertness and value representation, respectively. Additionally, the neural basis of moral ambivalence is associated with the TPJ, responsible for mentalizing.

An externally validated resting-state brain connectivity signature of pain-related learning

Pain can be conceptualized as a precision signal for reinforcement learning in the brain and alterations in these processes are a hallmark of chronic pain conditions. Investigating individual differences in pain-related learning therefore holds important clinical and translational relevance. Here, we developed and externally validated a novel resting-state brain connectivity-based predictive model of pain-related learning. The pre-registered external validation indicates that the proposed model explains 8-12% of the inter-individual variance in pain-related learning. Model predictions are driven by connections of the amygdala, posterior insula, sensorimotor, frontoparietal, and cerebellar regions, outlining a network commonly described in aversive learning and pain. We propose the resulting model as a robust and highly accessible biomarker candidate for clinical and translational pain research, with promising implications for personalized treatment approaches and with a high potential to advance our understanding of the neural mechanisms of pain-related learning.

Age-related differences in resting-state, task-related, and structural brain connectivity: graph theoretical analyses and visual search performance.

Previous magnetic resonance imaging (MRI) research suggests that aging is associated with a decrease in the functional interconnections within and between groups of locally organized brain regions (modules). Further, this age-related decrease in the segregation of modules appears to be more pronounced for a task, relative to a resting state, reflecting the integration of functional modules and attentional allocation necessary to support task performance. Here, using graph-theoretical analyses, we investigated age-related differences in a whole-brain measure of module connectivity, system segregation, for 68 healthy, community-dwelling individuals 18-78years of age. We obtained resting-state, task-related (visual search), and structural (diffusion-weighted) MRI data. Using a parcellation of modules derived from the participants' resting-state functional MRI data, we demonstrated that the decrease in system segregation from rest to task (i.e., reconfiguration) increased with age, suggesting an age-related increase in the integration of modules required by the attentional demands of visual search. Structural system segregation increased with age, reflecting weaker connectivity both within and between modules. Functional and structural system segregation had qualitatively different influences on age-related decline in visual search performance. Functional system segregation (and reconfiguration) influenced age-related decline in the rate of visual evidence accumulation (drift rate), whereas structural system segregation contributed to age-related slowing of encoding and response processes (nondecision time). The age-related differences in the functional system segregation measures, however, were relatively independent of those associated with structural connectivity.

Effect of positive social comparative feedback on the resting state connectivity of dopaminergic neural pathways: A preliminary investigation

Positive social comparative feedback is hypothesized to generate a dopamine response in the brain, similar to reward, by enhancing expectancies to support motor skill learning. However, no studies have utilized neuroimaging to examine this hypothesized dopaminergic mechanism. Therefore, the aim of this preliminary study was to investigate the effect of positive social comparative feedback on dopaminergic neural pathways measured by resting state connectivity. Thirty individuals practiced an implicit, motor sequence learning task and were assigned to groups that differed in feedback type. One group received feedback about their actual response time to complete the task (RT ONLY), while the other group received feedback about their response time with positive social comparison (RT + POS). Magnetic resonance imaging was acquired at the beginning and end of repetitive motor practice with feedback to measure practice-dependent changes in resting state brain connectivity. While both groups showed improvements in task performance and increases in performance expectancies, ventral tegmental area and the left nucleus accumbens (mesolimbic dopamine pathway) resting state connectivity increased in the RT + POS group but not in the RT ONLY group. Instead, the RT ONLY group showed increased connectivity between ventral tegmental area and primary motor cortex. Positive social comparative feedback during practice of a motor sequence task may induce a dopaminergic response in the brain along the mesolimbic pathway. However, given that absence of effects on expectancies and motor learning, more robust and individualized approaches may be needed to provide beneficial psychological and behavioral effects.

Fatigue and perceived energy in a sample of older adults over 10 years: A resting state functional connectivity study of neural correlates

PurposeDeclining energy and increasing fatigue, common in older age, predict neurodegenerative conditions, but their neural substrates are not known. We examined brain resting state connectivity in relation to declining self-reported energy levels (SEL) and occurrence of fatigue over time. MethodsWe examined resting-state functional MRI in 272 community dwelling older adults participating in the Health Aging and Body Composition Study (mean age 83 years; 57.4 % female; 40.8 % Black) with measures of fatigue and SEL collected at regular intervals over the prior ten years. Functional connectivity (FC) between cortex and striatum was examined separately for sensorimotor, executive, and limbic functional subregions. Logistic regression tested the association of FC in each network with prior fatigue state (reporting fatigue at least once or never reporting fatigue), and with SEL decline (divided into stable or declining SEL groups) and adjusted for demographic, physical function, mood, cognition, and comorbidities. ResultsHigher cortico-striatal FC in the right limbic network was associated with lower odds of reporting fatigue (better) at least once during the study period (adjusted odds ratio [95 % confidence interval], p-value: (0.747 [0.582, 0.955], 0.020), independent of SEL. Higher cortico-striatal FC in the right executive network was associated with higher odds of declining SEL (worse) during the study period (adjusted odds ratio [95 % confidence interval], p-value: (1.31 [1.01, 1.69], 0.041), independent of fatigue. Associations with other networks were not significant. ConclusionsIn this cohort of older adults, the cortico-striatal functional connectivity of declining SEL appears distinct from that underlying fatigue. Studies to further assess the neural correlates of energy and fatigue, and their independent contribution to neurodegenerative conditions are warranted.

Cue-exposure treatment influences resting-state functional connectivity—a randomized controlled fMRI study in alcohol use disorder

RationaleCue-exposure therapy (CET) consists of exposing patients to the cause of their affliction in a controlled environment and after psychological preparation. Ever since it was conceived, it has been suggested as a treatment for different types of behavioural impairments, from anxiety disorders to substance abuse. In the field of addictive behaviour, many different findings have been shown regarding the effectiveness of this therapy.ObjectivesThis study aims to examine the underlying neurobiological mechanisms of the effects of CET in patients with alcohol use disorder using resting-state functional magnetic resonance imaging (rs-fMRI).MethodsIn a randomized, controlled study, we examined patients after inpatient detoxification as well as healthy controls. Patients underwent nine sessions of CET spaced over 3 weeks. Rs-fMRI was conducted before treatment and 3 weeks after treatment onset in patients, healthy controls received only one rs-fMRI measurement. The final participant sample with complete data included 35 patients in the CET group, 17 patients in the treatment-as-usual group, and 43 HCs.ResultsOur results show differences in the Salience Network when comparing the CET group to the treatment-as-usual group (TAU). Functional connectivity between the anterior cingulate Cortex (ACC) and the insula was increased after CET, whereas it was decreased from ACC to the putamen and globus pallidus. Further, increased connectivity with the precuneus was found in the dorsal attention network after cue exposure treatment.ConclusionsThese findings suggest that cue exposure therapy changes the resting-state brain connectivity with additional effects to the standard psychotherapy treatment. Hence, our study results suggest why including CET in standard therapies might improve the preparation of patients in front of daily situations.

Exploring the neural and behavioral correlates of cognitive telerehabilitation in mild cognitive impairment with three distinct approaches.

Currently, the impact of drug therapies on neurodegenerative conditions is limited. Therefore, there is a strong clinical interest in non-pharmacological interventions aimed at preserving functionality, delaying disease progression, reducing disability, and improving quality of life for both patients and their caregivers. This longitudinal multicenter Randomized Controlled Trial (RCT) applies three innovative cognitive telerehabilitation (TR) methods to evaluate their impact on brain functional connectivity reconfigurations and on the overall level of cognitive and everyday functions. We will include 110 participants with mild cognitive impairment (MCI). Fifty-five participants will be randomly assigned to the intervention group who will receive cognitive TR via three approaches, namely: (a) Network-based Cognitive Training (NBCT), (b) Home-based Cognitive Rehabilitation (HomeCoRe), or (c) Semantic Memory Rehabilitation Training (SMRT). The control group (n = 55) will receive an unstructured home-based cognitive stimulation. The rehabilitative program will last either 4 (NBTC) or 6 weeks (HomeCoRe and SMRT), and the control condition will be adapted to each TR intervention. The effects of TR will be tested in terms of Δ connectivity change, obtained from high-density electroencephalogram (HD-EEG) or functional magnetic resonance imaging at rest (rs-fMRI), acquired before (T0) and after (T1) the intervention. All participants will undergo a comprehensive neuropsychological assessment at four time-points: baseline (T0), within 2 weeks (T1), and after 6 (T2) and 12 months (T3) from the end of TR. The results of this RCT will identify a potential association between improvement in performance induced by individual cognitive TR approaches and modulation of resting-state brain connectivity. The knowledge gained with this study might foster the development of novel TR approaches underpinned by established neural mechanisms to be validated and implemented in clinical practice.Clinical trial registration: [https://classic.clinicaltrials.gov/ct2/show/NCT06278818], identifier [NCT06278818].

Lack of reproducibility of resting-state functional MRI findings in migraine with aura.

Several studies have applied resting-state functional MRI to examine whether functional brain connectivity is altered in migraine with aura patients. These studies had multiple limitations, including small sample sizes, and reported conflicting results. Here, we performed a large, cross-sectional brain imaging study to reproduce previous findings. We recruited women aged 30-60 years from the nationwide Danish Twin Registry. Resting-state functional MRI of women with migraine with aura, their co-twins, and unrelated migraine-free twins was performed at a single centre. We carried out an extensive series of brain connectivity data analyses. Patients were compared to migraine-free controls and to co-twins. Comparisons were based on data from 160 patients, 30 co-twins, and 136 controls. Patients were similar to controls with regard to age, and several lifestyle characteristics. We replicated clear effects of age on resting-state networks. In contrast, we failed to detect any differences, and to replicate previously reported differences, in functional connectivity between migraine patients with aura and non-migraine controls or their co-twins in any of the analyses. Given the large sample size and the unbiased population-based design of our study, we conclude that women with migraine with aura have normal resting-state brain connectivity outside of migraine attacks.

Fundamental behavioral and neurophysiologic relationships between sensory processing, intolerance of uncertainty, and autistic traits in children: A hybrid approach

Sensory differences are common and often challenging for autistic children. Furthermore, atypical sensory processing is associated with autistic traits and other autism-related behaviors, such as intolerance of uncertainty (IU). Such traits and their relatedness vary continuously across autistic and non-autistic children alike. However, the underlying neural correlates of these continuous variables, and their associations, are not well understood. Therefore, this study examined relationships between sensory processing, IU, autistic traits, and associated resting state brain connectivity, across a sample of both autistic (n = 30) and non-autistic (n = 26) children. In addition to computing behavioral correlations between these factors, we carried out independent component network functional connectivity analysis to investigate associations between cortical and cerebellar networks and behavioral results between groups and across our entire sample. Across-group correlations between sensory processing, autistic traits, and IU were significant. In addition, data demonstrated overlapping sensory processing and intolerance of uncertainty scores, spanning the groups. Brain (rs-fMRI)—behavioral relationships revealed strong associations between sensory, large-scale resting state, and cerebellar networks and behavioral scores. Overall, our findings suggest that sensory differences are related to IU and autistic traits across the population. Neurophysiologic data pointed to functional connectivity between sensory cortices and supramodal brain networks. These findings provide evidence for the continuous variation of behaviors common to autism throughout the entire population and their neurobiological correlates.

Resting-state brain connectivity correlates of musical sophistication.

A growing body of research has investigated how performing arts training, and more specifically, music training, impacts the brain. Recent meta-analytic work has identified multiple brain areas where activity varies as a function of levels of musical expertise gained through music training. However, research has also shown that musical sophistication may be high even without music training. Thus, we aim to extend previous work by investigating whether the functional connectivity of these areas relates to interindividual differences in musical sophistication, and to characterize differences in connectivity attributed to performing arts training. We analyzed resting-state functional magnetic resonance imaging from n = 74 participants, of whom 37 received performing arts training, that is, including a musical instrument, singing, and/or acting, at university level. We used a validated, continuous measure of musical sophistication to further characterize our sample. Following standard pre-processing, fifteen brain areas were identified a priori based on meta-analytic work and used as seeds in separate seed-to-voxel analyses to examine the effect of musical sophistication across the sample, and between-group analyses to examine the effects of performing arts training. Connectivity of bilateral superior temporal gyrus, bilateral precentral gyrus and cerebellum, and bilateral putamen, left insula, and left thalamus varied with different aspects of musical sophistication. By including these measures of these aspects as covariates in post hoc analyses, we found that connectivity of the right superior temporal gyrus and left precentral gyrus relate to effects of performing arts training beyond effects of individual musical sophistication. Our results highlight the potential role of sensory areas in active engagement with music, the potential role of motor areas in emotion processing, and the potential role of connectivity between putamen and lingual gyrus in general musical sophistication.

Resting-state fractal brain connectivity is associated with impaired cognitive performance in healthy aging

Aging affects cognitive functions even in the absence of ongoing pathologies. The neurophysiological basis of age-related cognitive decline (CD), however, is not completely understood. Alterations in both functional brain connectivity and in the fractal scaling of neuronal dynamics have been linked to aging and cognitive performance. Recently, fractal connectivity (FrC) has been proposed — combining the two concepts — for capturing long-term interactions among brain regions. FrC was shown to be influenced by increased mental workload; however, no prior studies investigated how resting-state FrC relates to cognitive performance and plausible CD in healthy aging. We recruited 19 healthy elderly (HE) and 24 young control (YC) participants, who underwent resting-state electroencephalography (EEG) measurements and comprehensive cognitive evaluation using 7 tests of the Cambridge Neurophysiological Test Automated Battery. FrC networks were reconstructed from EEG data using the recently introduced multiple-resampling cross-spectral analysis (MRCSA). Elderly individuals could be characterized with increased response latency and reduced performance in 4–4 tasks, respectively, with both reaction time and accuracy being affected in two tasks. Auto- and cross-spectral exponents — characterizing regional fractal dynamics and FrC, respectively, — were found reduced in HE when compared to YC over most of the cortex. Additionally, fractal scaling of frontoparietal connections expressed an inverse relationship with task performance in visual memory and sustained attention domains in elderly, but not in young individuals. Our results confirm that the fractal nature of brain connectivity — as captured by MRCSA — is affected in healthy aging. Furthermore, FrC appears as a sensitive neurophysiological marker of age-related CD.

Metreleptin Robustly Increases Resting-state Brain Connectivity in Treatment-naïve Female Patients With Lipodystrophy.

Research in lipodystrophy (LD) and its treatment with metreleptin has not only helped patients with LD but has opened new directions in investigating leptin's role in metabolism and the regulation of eating behavior. Previously, in a study with patients with LD undergoing metreleptin treatment using functional magnetic resonance imaging (MRI), we found significantly increased resting-state brain connectivity in 3 brain areas including the hypothalamus. In this study, we aimed to reproduce our functional MRI findings in an independent sample and compare results to healthy participants. Measurements in 4 female patients with LD undergoing metreleptin treatment and 3 healthy untreated controls were performed at 4 different time points over 12 weeks. To identify treatment-related brain connectivity alterations, eigenvector centrality was computed from resting-state functional MRI data for each patient and each session. Thereafter, analysis aimed at detecting consistent brain connectivity changes over time across all patients. In parallel to metreleptin treatment of the patients with LD, we found a significant brain connectivity increase in the hypothalamus and bilaterally in posterior cingulate gyrus. Using a 3-factorial model, a significant interaction between group and time was found in the hypothalamus. Investigating brain connectivity alterations with metreleptin treatment using an independent sample of patients with LD, we have reproduced an increase of brain connectivity in hedonic and homeostatic central nervous networks observed previously with metreleptin treatment. These results are an important contribution to ascertain brain leptin action and help build a foundation for further research of central nervous effects of this important metabolic hormone.

Predicting Alcohol-Related Memory Problems in Older Adults: A Machine Learning Study with Multi-Domain Features.

Memory problems are common among older adults with a history of alcohol use disorder (AUD). Employing a machine learning framework, the current study investigates the use of multi-domain features to classify individuals with and without alcohol-induced memory problems. A group of 94 individuals (ages 50-81 years) with alcohol-induced memory problems (the memory group) were compared with a matched control group who did not have memory problems. The random forests model identified specific features from each domain that contributed to the classification of the memory group vs. the control group (AUC = 88.29%). Specifically, individuals from the memory group manifested a predominant pattern of hyperconnectivity across the default mode network regions except for some connections involving the anterior cingulate cortex, which were predominantly hypoconnected. Other significant contributing features were: (i) polygenic risk scores for AUD, (ii) alcohol consumption and related health consequences during the past five years, such as health problems, past negative experiences, withdrawal symptoms, and the largest number of drinks in a day during the past twelve months, and (iii) elevated neuroticism and increased harm avoidance, and fewer positive "uplift" life events. At the neural systems level, hyperconnectivity across the default mode network regions, including the connections across the hippocampal hub regions, in individuals with memory problems may indicate dysregulation in neural information processing. Overall, the study outlines the importance of utilizing multidomain features, consisting of resting-state brain connectivity data collected ~18 years ago, together with personality, life experiences, polygenic risk, and alcohol consumption and related consequences, to predict the alcohol-related memory problems that arise in later life.

N°247 – Resting-state brain connectivity differences between epileptic and non-epileptic first events

N°247 – Resting-state brain connectivity differences between epileptic and non-epileptic first events

Neural correlates of Type A personality: Type A personality mediates the association of resting-state brain activity and connectivity with eating disorder symptoms

BackgroundType A personality (TAP) was characterized by impatience, competitiveness, aggressiveness, and hostility. Higher TAP was proved to be associated with more eating disorder symptoms (EDS). While little is known about the underlying neural substrates of TAP and how TAP is linked to EDS at the neural level. MethodsTo investigate the neural basis of TAP, we adopted fractional amplitude of low-frequency fluctuations (fALFF) and resting-state functional connectivity (RSFC) via resting-state functional magnetic resonance imaging (rs-fMRI) (N = 1620). Mediation models were examined to explore the relationship between TAP, EDS, and brain activity. ResultsTAP was associated with decreased fALFF in the left middle frontal gyrus (MFG) and increased fALFF in the left precentral gyrus (PreCG). Furthermore, TAP was positively correlated to RSFC between the left MFG and left inferior temporal gyrus (ITG) and between the left PreCG and right middle temporal gyrus (MTG). Mediation analysis showed TAP fully mediated the association of the left MFG activity, MFG-ITG connectivity, and PreCG-MTG connectivity with EDS. LimitationsThe cross-sectional design of this study precludes us from specifying the causal relationship in the associations we observed. ConclusionsOur results suggested spontaneous activity in the left MFG and PreCG is associated with TAP, and even in general sample, people with higher TAP showed more EDS. The present study is the first to investigate the neurobiological underpinnings of TAP in a large sample and further offered new insights into the relation between TAP and EDS from a neural basis perspective.

High and low worriers do not differ in unstimulated resting-state brain connectivity

Chronic, excessive and uncontrollable worry presents an anxiety rising and distressing mental activity relevant in a range of psychological disorders. Task based studies investigating its underlying neural mechanisms reveal fairly heterogenous results. The current study aimed to investigate pathological worry related effects on the functional neural network architecture in the resting unstimulated brain. Using resting-state functional magnetic resonance imaging (rsfMRI) we compared functional connectivity (FC) patterns between 21 high worriers and 21 low worriers. We, on the one hand, conducted a seed-to-voxel analysis based on recent meta-analytic findings and, on the other hand, implemented a data-driven multi voxel pattern analysis (MVPA) approach to yield brain clusters showing connectivity differences between the two groups. Additionally, the seed regions and MVPA were used to investigate whether whole brain connectivity is associated with momentary state worry across groups. The data did not reveal differences in resting-state FC related to pathological worry, neither by the seed-to-voxel or MVPA approach testing for differences linked to trait worry nor by using the MVPA to test for state worry related aberrations. We discuss whether the null findings in our analyses are related to spontaneous fluctuations in momentary worry and the associated presence of multiple fluctuating brain states that could cause mutually cancelling effects. For future studies investigating the neural correlates of excessive worry, we propose a direct worry induction for better control of the situation.

Functional connectivity signatures of major depressive disorder: machine learning analysis of two multicenter neuroimaging studies

The promise of machine learning has fueled the hope for developing diagnostic tools for psychiatry. Initial studies showed high accuracy for the identification of major depressive disorder (MDD) with resting-state connectivity, but progress has been hampered by the absence of large datasets. Here we used regular machine learning and advanced deep learning algorithms to differentiate patients with MDD from healthy controls and identify neurophysiological signatures of depression in two of the largest resting-state datasets for MDD. We obtained resting-state functional magnetic resonance imaging data from the REST-meta-MDD (N = 2338) and PsyMRI (N = 1039) consortia. Classification of functional connectivity matrices was done using support vector machines (SVM) and graph convolutional neural networks (GCN), and performance was evaluated using 5-fold cross-validation. Features were visualized using GCN-Explainer, an ablation study and univariate t-testing. The results showed a mean classification accuracy of 61% for MDD versus controls. Mean accuracy for classifying (non-)medicated subgroups was 62%. Sex classification accuracy was substantially better across datasets (73–81%). Visualization of the results showed that classifications were driven by stronger thalamic connections in both datasets, while nearly all other connections were weaker with small univariate effect sizes. These results suggest that whole brain resting-state connectivity is a reliable though poor biomarker for MDD, presumably due to disease heterogeneity as further supported by the higher accuracy for sex classification using the same methods. Deep learning revealed thalamic hyperconnectivity as a prominent neurophysiological signature of depression in both multicenter studies, which may guide the development of biomarkers in future studies.

A machine learning based approach towards high-dimensional mediation analysis

Mediation analysis is used to investigate the role of intermediate variables (mediators) that lie in the path between an exposure and an outcome variable. While significant research has focused on developing methods for assessing the influence of mediators on the exposure-outcome relationship, current approaches do not easily extend to settings where the mediator is high-dimensional. These situations are becoming increasingly common with the rapid increase of new applications measuring massive numbers of variables, including brain imaging, genomics, and metabolomics. In this work, we introduce a novel machine learning based method for identifying high dimensional mediators. The proposed algorithm iterates between using a machine learning model to map the high-dimensional mediators onto a lower-dimensional space, and using the predicted values as input in a standard three-variable mediation model. Hence, the machine learning model is trained to maximize the likelihood of the mediation model. Importantly, the proposed algorithm is agnostic to the machine learning model that is used, providing significant flexibility in the types of situations where it can be used. We illustrate the proposed methodology using data from two functional Magnetic Resonance Imaging (fMRI) studies. First, using data from a task-based fMRI study of thermal pain, we combine the proposed algorithm with a deep learning model to detect distributed, network-level brain patterns mediating the relationship between stimulus intensity (temperature) and reported pain at the single trial level. Second, using resting-state fMRI data from the Human Connectome Project, we combine the proposed algorithm with a connectome-based predictive modeling approach to determine brain functional connectivity measures that mediate the relationship between fluid intelligence and working memory accuracy. In both cases, our multivariate mediation model links exposure variables (thermal pain or fluid intelligence), high dimensional brain measures (single-trial brain activation maps or resting-state brain connectivity) and behavioral outcomes (pain report or working memory accuracy) into a single unified model. Using the proposed approach, we are able to identify brain-based measures that simultaneously encode the exposure variable and correlate with the behavioral outcome.

Early screening model for mild cognitive impairment based on resting-state functional connectivity: a functional near-infrared spectroscopy study.

As an early stage of Alzheimer's disease (AD), the diagnosis of amnestic mild cognitive impairment (aMCI) has important clinical value for timely intervention of AD. Functional near-infrared spectroscopy (fNIRS)-based resting-state brain connectivity analysis, which could provide an economic and quick screening strategy for aMCI, remains to be extensively investigated. This study aimed to verify the feasibility of fNIRS-based resting-state brain connectivity for evaluating brain function in patients with aMCI, and to determine an early screening model for auxiliary diagnosis. The resting-state fNIRS was utilized for exploring the changes in functional connectivity of 64 patients with aMCI. The region of interest (ROI)-based and channel-based connections with significant inter-group differences have been extracted through the two-sample -tests and the receiver operating characteristic (ROC). These connections with specificity and sensitivity were then taken as features for classification. Compared with healthy controls, connections of the MCI group were significantly reduced between the bilateral prefrontal, parietal, occipital, and right temporal lobes. Specifically, the long-range connections from prefrontal to occipital lobe, and from prefrontal to parietal lobe, exhibited stronger identifiability (area under the ROC curve , ** ). Subsequently, the optimal classification accuracy of ROI-based connections was 71.59%. Furthermore, the most responsive connections were located between the right dorsolateral prefrontal lobe and the left occipital lobe, concomitant with the highest classification accuracy of 73.86%. Our findings indicate that fNIRS-based resting-state functional connectivity analysis could support MCI diagnosis. Notably, long-range connections involving the prefrontal and occipital lobes have the potential to be efficient biomarkers.