Brain Functional Connectivity Measures Research Articles

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.

The brain functional connectivity in the default mode network is associated with self-efficacy in young adults.

Self-efficacy is a subjective belief that depends on self-related past experience, and is a strong predictor for individual future performance. To the aim of promoting one's future performance, it is necessary to gain better knowledge of it's cognitive process and brain mechanism. The present research sought to investigate the functional connectivity basis of self-efficacy by using the resting-state functional magnetic resonance imaging data of a large sample of young adults (536 participants). Multiple regression analysis was performed to examine the relationship between self-efficacy integrated score and brain functional connectivity measures. Gender, age, mean framewise displacement and grey matter volume were used as nuisance covariates. The whole-brain analysis revealed an association between self-efficacy and the functional connectivity of several regions within the default mode network. These regions included the right anterior cingulate cortex, the left posterior cingulate cortex/precuneus and bilateral parahippocampal cortex. Our findings suggest that the default mode network plays a crucial role in self-efficacy, and hold the view that episodic memory and self-related processing have influence on self-efficacy.

Regional Disconnection in Alzheimer Dementia and Amyloid-Positive Mild Cognitive Impairment: Association Between EEG Functional Connectivity and Brain Glucose Metabolism.

Introduction: The disconnection hypothesis of Alzheimer's disease (AD) is supported by growing neuroimaging and neurophysiological evidence of altered brain functional connectivity in cognitively impaired individuals. Brain functional modalities such as [18F]fluorodeoxyglucose positron-emission tomography ([18F]FDG-PET) and electroencephalography (EEG) measure different aspects of synaptic functioning, and can contribute to understanding brain connectivity disruptions in AD.Aim: This study investigated the relationship between cortical glucose metabolism and topographical EEG measures of brain functional connectivity in subjects along AD continuum.Methods: Patients diagnosed with mild cognitive impairment (MCI) and AD (n = 67), and stratified into amyloid-positive (n = 32) and negative (n = 10) groups according to cerebrospinal fluid Aβ42/40 ratio, were assessed with [18F]FDG-PET and resting-state EEG recordings. EEG-based neuroimaging analysis involved standardized low-resolution electromagnetic tomography (sLORETA), which estimates functional connectivity from cortical sources of electrical activity in a 3D head model.Results: Glucose hypometabolism in temporoparietal lobes was significantly associated with altered EEG functional connectivity of the same regions of interest in clinically diagnosed MCI and AD patients and in patients with biomarker-verified AD pathology. The correlative pattern of disrupted connectivity in temporoparietal lobes, as detected by EEG sLORETA analysis, included decreased instantaneous linear connectivity in fast frequencies and increased lagged linear connectivity in slow frequencies in relation to the activity of remaining cortex.Conclusions: Topographical EEG measures of functional connectivity detect regional dysfunction of AD-vulnerable brain areas as evidenced by association and spatial overlap with the cortical glucose hypometabolism in MCI and AD patients.Impact statementThe association between glucose hypometabolism, as evidenced by [18F]FDG-PET ([18F]fluorodeoxyglucose positron-emission tomography), and altered electroencephalography (EEG) functional connectivity metrics within temporoparietal lobes provides link between synaptic, neurophysiological, and metabolic impairment in mild cognitive impairment and Alzheimer's disease patients. This study reported alterations in EEG measures of both instantaneous and lagged linear connectivity across distinct frequency bands, both of which were shown to be important for inter- and intrahemispheric communication and function of memory systems in general. EEG-based imaging of brain functional connectivity has a potential to serve as a noninvasive, low-cost, and widely available alternative in assessing synaptic and network dysfunction in cognitively impaired patients.

Identifying commonality and specificity across psychosis sub-groups via classification based on features from dynamic connectivity analysis

It is difficult to distinguish schizophrenia (SZ), schizoaffective disorder (SAD), and bipolar disorder with psychosis (BPP) as their clinical diagnoses rely on symptoms that overlap. In this paper, we investigate if there is biological evidence to support the symptom-based clinical categories by looking across the three disorders using dynamic connectivity measures, and provide meaningful characteristics on which brain functional connectivity measures are commonly or uniquely impaired. Large-sample functional magnetic resonance image (fMRI) datasets from 623 subjects including 238 healthy controls (HCs), 113 SZ patients, 132 SAD patients, and 140 BPP patients were analyzed. First, we computed whole-brain dynamic functional connectivity (DFC) using a sliding-window technique, and then extracted the individual connectivity states by applying our previously proposed decomposition-based DFC analysis method. Next, with the features from the dominant connectivity state, we assessed the clinical categories by performing both four-group (SZ, SAD, BPP and healthy control groups) and pair-wise classification using a support vector machine within cross-validation. Furthermore, we comprehensively summarized the shared and unique connectivity alterations among the disorders. In terms of the classification performance, our method achieved 69% in the four-group classification and >80% in the between-group classifications for the mean overall accuracy; and yielded 66% in the four-group classification and >80% in the between-group classifications for the mean balanced accuracy. Through summarizing the features that were automatically selected in the classifications, we found that among the three symptom-related disorders, their disorder-common impairments primarily included the decreased connectivity strength between thalamus and cerebellum and the increased strength between postcentral gyrus and thalamus. The disorder-unique changes included more various brain regions, mainly in the temporal and frontal gyrus. Our work demonstrates that dynamic functional connectivity provides biological evidence that both common and unique impairments exist in psychosis sub-groups.

Characterizing impulsivity and resting-state functional connectivity in normal-weight binge eaters.

Binge eating is characterized by episodes of uncontrolled eating, within discrete periods of time. Although it is usually described in obese individuals or as a symptom of Binge Eating Disorder (BED), this behavior can also occur in the normal-weight (NW) population. An interesting premise suggests that impulsivity might contribute to the onset of binge eating and the progression toward weight gain. Drawing upon this evidence, here we explored impulsivity in NW individuals reporting binge-eating episodes through a functional connectivity approach. We hypothesized that, even in the absence of an eating disorder, NW binge eaters would be characterized by connectivity pattern changes in corticostriatal regions implicated in impulsivity, similarly to the results described in BED individuals. A resting-state functional magnetic resonance imaging study tested 39 NW men and women, with and without binge eating (binge eaters, BE and non-BE). Brain functional connectivity was explored by means of graph theoretic centrality measures and traditional seed-based analysis; trait impulsivity was assessed with self-report questionnaires. The BE group was characterized by a higher degree of trait impulsivity. Brain functional connectivity measures revealed lower degree centrality within the right middle frontal gyrus, left insula/putamen and left temporoparietal regions and a lower functional connectivity between the right middle frontal gyrus and right insula in the BE group. The results support previous evidence on BED of altered functional connectivity and higher impulsivity at the roots of overeating behavior, but further extend this concept excluding any potential confounding effect exerted by the weight status.

Neural Correlates of Abnormal Temporal Discrimination in Unaffected Relatives of Cervical Dystonia Patients.

Background: An abnormal temporal discrimination threshold in cervical dystonia (CD) is considered to be a mediational endophenotype; in unaffected relatives it is hypothesized to indicate non-manifesting gene carriage. The pathogenesis underlying this condition remains unknown. Investigation of the neural networks involved in disordered temporal discrimination may highlight its pathomechanisms.Objective: To examine resting state brain function in unaffected relatives of CD patients with normal and abnormal temporal discrimination. We hypothesized that the endophenotype, an abnormal temporal discrimination, would manifest as altered connectivity in relatives in regions associated with CD, thereby illuminating the neural substrates of the link between temporal discrimination and CD.Methods: Rs-fMRI data was analyzed from two sex- and age-matched cohorts: 16 unaffected relatives of CD patients with normal temporal discrimination and 16 with abnormal temporal discrimination. Regional and whole brain functional connectivity measures were extracted via Independent Component Analysis (ICA), Regional Homogeneity (ReHo), and Amplitude of Low Frequency (ALFF) analyses.Results: Our ICA analysis revealed increased connectivity within both the executive control and cerebellar networks and decreased connectivity within the sensorimotor network in relatives with abnormal temporal discrimination when compared to relatives with normal temporal discrimination. The ReHo and ALFF analyses complimented these results and demonstrated connectivity differences in areas corresponding to motor planning, movement coordination, visual information processing, and eye movements in unaffected relatives with abnormal temporal discrimination.Conclusion: Disordered connectivity in unaffected relatives with abnormal temporal discrimination illuminates neural substrates underlying endophenotype expression and supports the hypothesis that genetically determined aberrant connectivity, when later coupled with unknown environmental triggers, may lead to disease penetrance.

Classification and Prediction of Brain Disorders Using Functional Connectivity: Promising but Challenging.

Brain functional imaging data, especially functional magnetic resonance imaging (fMRI) data, have been employed to reflect functional integration of the brain. Alteration in brain functional connectivity (FC) is expected to provide potential biomarkers for classifying or predicting brain disorders. In this paper, we present a comprehensive review in order to provide guidance about the available brain FC measures and typical classification strategies. We survey the state-of-the-art FC analysis methods including widely used static functional connectivity (SFC) and more recently proposed dynamic functional connectivity (DFC). Temporal correlations among regions of interest (ROIs), data-driven spatial network and functional network connectivity (FNC) are often computed to reflect SFC from different angles. SFC can be extended to DFC using a sliding-window framework, and intrinsic connectivity states along the time-varying connectivity patterns are typically extracted using clustering or decomposition approaches. We also briefly summarize window-less DFC approaches. Subsequently, we highlight various strategies for feature selection including the filter, wrapper and embedded methods. In terms of model building, we include traditional classifiers as well as more recently applied deep learning methods. Moreover, we review representative applications with remarkable classification accuracy for psychosis and mood disorders, neurodevelopmental disorder, and neurological disorders using fMRI data. Schizophrenia, bipolar disorder, autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD), Alzheimer's disease and mild cognitive impairment (MCI) are discussed. Finally, challenges in the field are pointed out with respect to the inaccurate diagnosis labeling, the abundant number of possible features and the difficulty in validation. Some suggestions for future work are also provided.

Functional connectivity pattern during motor imagery and execution of paretic hand movements of stroke patients

Motor imagery (MI) has emerged as a promising technique for promoting neuroplasticity and recovery of motor function after a stroke. It is used as an alternative to movement execution (ME). The brain structures involved in MI are largely documented but the existent functional connectivity (FC) during MI and the differences with ME in stroke patients remain underinvestigated. Functional magnetic resonance imaging (fMRI) techniques have the potential to reveal these changes. We aimed to investigate the FC during MI and ME and correlate FC measures between sensorimotor (SMN) regions with motor function of stroke patients. Twenty-four stroke patients with hemiparesis (time post-stroke = 10 ± 21 months; Upper-limb Fugl-Meyer Score (FMS) = 27.2 ± 13.5/66) were included. The patients underwent fMRI examination while they either imagined or executed wrist extension with their affected hand (AH). We studied FC using seed-to-voxels analysis. We compared the FC of SMN regions between MI and ME of AH movement. Additionally, correlations among brain FC measures and behavioral scores of motor function (FMS) were computed. During MI of the AH in comparison with ME, FC of the ipsilesional SMN (BA 40) was decreased with the ipsilesional somatosensory cortex (BA 2, 40, 5) and the cerebellum. In parallel, it was increased with the bilateral frontal cortex (BA 8, 10) ( Fig. 1 ). Besides this, correlation analyses with the FMS revealed that during MI, more impaired patients relied more on functional connections between contralasional BA4 and cerebellum (crus 1 and 2) than less impaired patients did. The connectivity changes observed may reveal compensatory networks for MI implicating more the frontal lobe and the contralesional hemisphere in most impaired patients. These results characterized the connectivity patterns of ME and MI and may provide new insights into the neural mechanism underlying them.

Human fronto-parietal response scattering subserves vigilance at night

Lack of sleep has a considerable impact on vigilance: we perform worse, we make more errors, particularly at night, when we should be sleeping. Measures of brain functional connectivity suggest that decrease in vigilance during sleep loss is associated with an impaired cross-talk within the fronto-parietal cortex. However, fronto-parietal effective connectivity, which is more closely related to the causal cross-talk between brain regions, remains unexplored during prolonged wakefulness. In addition, no study has simultaneously investigated brain effective connectivity and wake-related changes in vigilance, preventing the concurrent incorporation of the two aspects. Here, we used electroencephalography (EEG) to record responses evoked by Transcranial Magnetic Stimulation (TMS) applied over the frontal lobe in 23 healthy young men (18–30 yr.), while they simultaneously performed a vigilance task, during 8 sessions spread over 29 h of sustained wakefulness. We assessed Response Scattering (ReSc), an estimate of effective connectivity, as the propagation of TMS-evoked EEG responses over the fronto-parietal cortex. Results disclose a significant change in fronto-parietal ReSc with time spent awake. When focusing on the night-time period, when one should be sleeping, participants with lower fronto-parietal ReSc performed worse on the vigilance task. Conversely, no association was detected during the well-rested, daytime period. Night-time fronto-parietal ReSc also correlated with objective EEG measures of sleepiness and alertness. These changes were not accompanied by variations in fronto-parietal response complexity. These results suggest that decreased brain response propagation within the fronto-parietal cortex is associated to increased vigilance failure during night-time prolonged wakefulness. This study reveals a novel facet of the detrimental effect on brain function of extended night-time waking hours, which is increasingly common in our societies.

Resting-state functional connectivity predicts longitudinal pain symptom change in urologic chronic pelvic pain syndrome: a MAPP network study.

Chronic pain symptoms often change over time, even in individuals who have had symptoms for years. Studying biological factors that predict trends in symptom change in chronic pain may uncover novel pathophysiological mechanisms and potential therapeutic targets. In this study, we investigated whether brain functional connectivity measures obtained from resting-state functional magnetic resonance imaging at baseline can predict longitudinal symptom change (3, 6, and 12 months after scan) in urologic chronic pelvic pain syndrome. We studied 52 individuals with urologic chronic pelvic pain syndrome (34 women, 18 men) who had baseline neuroimaging followed by symptom tracking every 2 weeks for 1 year as part of the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Research Network study. We found that brain functional connectivity can make a significant prediction of short-term (3 month) pain reduction with 73.1% accuracy (69.2% sensitivity and 75.0% precision). In addition, we found that the brain regions with greatest contribution to the classification were preferentially aligned with the left frontoparietal network. Resting-state functional magnetic resonance imaging measures seemed to be less informative about 6- or 12-month symptom change. Our study provides the first evidence that future trends in symptom change in patients in a state of chronic pain may be linked to functional connectivity within specific brain networks.

Adaptive testing for multiple traits in a proportional odds model with applications to detect SNP-brain network associations.

There has been increasing interest in developing more powerful and flexible statistical tests to detect genetic associations with multiple traits, as arising from neuroimaging genetic studies. Most of existing methods treat a single trait or multiple traits as response while treating an SNP as a predictor coded under an additive inheritance mode. In this paper, we follow an earlier approach in treating an SNP as an ordinal response while treating traits as predictors in a proportional odds model (POM). In this way, it is not only easier to handle mixed types of traits, e.g., some quantitative and some binary, but it is also potentially more robust to the commonly adopted additive inheritance mode. More importantly, we develop an adaptive test in a POM so that it can maintain high power across many possible situations. Compared to the existing methods treating multiple traits as responses, e.g., in a generalized estimating equation (GEE) approach, the proposed method can be applied to a high dimensional setting where the number of phenotypes (p) can be larger than the sample size (n), in addition to a usual small P setting. The promising performance of the proposed method was demonstrated with applications to the Alzheimer's Disease Neuroimaging Initiative (ADNI) data, in which either structural MRI driven phenotypes or resting-state functional MRI (rs-fMRI) derived brain functional connectivity measures were used as phenotypes. The applications led to the identification of several top SNPs of biological interest. Furthermore, simulation studies showed competitive performance of the new method, especially for p>n.

Alterations in dorsal and ventral posterior cingulate connectivity in APOE ε4 carriers at risk of Alzheimer's disease.

BackgroundRecent evidence suggests that exercise plays a role in cognition and that the posterior cingulate cortex (PCC) can be divided into dorsal and ventral subregions based on distinct connectivity patterns.AimsTo examine the effect of physical activity and division of the PCC on brain functional connectivity measures in subjective memory complainers (SMC) carrying the epsilon 4 allele of apolipoprotein E (APOE ε4) allele.MethodParticipants were 22 SMC carrying the APOE ε4 allele (ε4+; mean age 72.18 years) and 58 SMC non-carriers (ε4–; mean age 72.79 years). Connectivity of four dorsal and ventral seeds was examined. Relationships between PCC connectivity and physical activity measures were explored.Resultsε4+ individuals showed increased connectivity between the dorsal PCC and dorsolateral prefrontal cortex, and the ventral PCC and supplementary motor area (SMA). Greater levels of physical activity correlated with the magnitude of ventral PCC–SMA connectivity.ConclusionsThe results provide the first evidence that ε4+ individuals at increased risk of cognitive decline show distinct alterations in dorsal and ventral PCC functional connectivity.Declaration of interestD.A. has served on scientific advisory boards for Novartis, Eli Lilly, Janssen, Prana and Pfizer, and as Editor-in-Chief for International Psychogeriatrics; received speaker honoraria from Pfizer and Lundbeck, and research support from Eli Lilly, GlaxoSmithKline, Forest Laboratories, Novartis, and CSIRO. C.L.M. has received consulting fees from Eli Lilly and Prana Biotechnology, and has stock ownership in Prana Biotechnology. C.C.R. has received consultancy payments from Roche and Piramal, and research support from Avid Radiopharmaceuticals, Eli Lilly, GE Healthcare, Piramal and Navidea for amyloid imaging. C.S. has provided clinical consultancy and been on scientific advisory committees for the Australian CSIRO, Alzheimer's Australia, University of Melbourne and other relationships, which are subject to confidentiality clauses; she has been a named Chief Investigator on investigator-driven collaborative research projects in partnership with Pfizer, Merck, Piramal, Bayer and GE Healthcare. Her research programme has received support from the National Health and Medical Research Council Alzheimer's Association, Collier Trust, Scobie and Claire McKinnon Foundation, JO and JR Wicking Trust, Shepherd Foundation, Brain Foundation, Mason Foundation, Ramaciotti Foundation, Alzheimer's Australia and the Royal Australian College of Physicians.Copyright and usage© The Royal College of Psychiatrists 2015. This is an open access article distributed under the terms of the Creative Commons Non-Commercial, No Derivatives (CC BY-NC-ND) licence.

Abstract TP31: Resting State Functional MR Connectivity In Acute And Sub-acute Stroke

Objective: Stroke can have important effects on widespread brain regions resulting in network disruption. This study investigates the changes in spontaneous activity in the brain after stroke using resting-state functional connectivity(FC) MRI. Methods: Acute ischemic stroke patients (N=22, 11 cortical, mean age=60, 10F) were recruited within 7 days of stroke onset(timepoint 1, V1). Eleven of these patients were also scanned at timepoint 2(V2), approximately 3 months later. Age-matched healthy controls (N=17, mean age = 56, 7F) and 7 patients with risk factors for stroke (mean age=67, 1F) were also recruited in the study. Ten minute eyes-closed resting-state fMRI scans were collected along with a high resolution anatomical scan. We examined FC in the language network consisting of ten regions extracted based on functional activations on a phonemic fluency task performed during a separate scan. We also examined the correlation of brain FC with behavioral performance on the task outside the scanner. Results: We examined correlation between every seed region and every other region in the network(45 seed region-pairs). Compared to age-matched controls, acute strokes showed significantly reduced inter-hemispheric connectivity, specifically between inferior frontal and temporal regions(p<.001 corrected). This difference however was not significant in the sub-acute stage(Figure 1 left panel). Additionally, although both acute and sub-acute patients showed significant differences in performance compared to normals on phonemic fluency(Fig 1 right), there were no significant correlations between brain FC measures and behavior and no differences in the correlation between brain FC and behavior among the groups. Conclusions: These results suggest that acute strokes show reduced brain connectivity between contralateral regions in the language network; however, this connectivity pattern is restored in the sub-acute stage.

Brain functional connectivity in stimulant drug dependence and obsessive–compulsive disorder

There are reasons for thinking that obsessive-compulsive disorder (OCD) and drug dependence, although conventionally distinct diagnostic categories, might share important cognitive and neurobiological substrates. We tested this hypothesis directly by comparing brain functional connectivity measures between patients with OCD, stimulant dependent individuals (SDIs; many of whom were non-dependent users of other recreational drugs) and healthy volunteers. We measured functional connectivity between each possible pair of 506 brain regional functional MRI time series representing low frequency (0.03-0.06 Hz) spontaneous brain hemodynamics in healthy volunteers (N=18), patients with OCD (N=18) and SDIs (N=18). We used permutation tests to identify i) brain regions where strength of connectivity was significantly different in both patient groups compared to healthy volunteers; and ii) brain regions and connections which had significantly different functional connectivity between patient groups. We found that functional connectivity of right inferior and superior orbitofrontal cortex (OFC) was abnormally reduced in both disorders. Whether diagnosed as OCD or SDI, patients with higher scores on measures of compulsive symptom severity showed greater reductions of right orbitofrontal connectivity. Functional connections specifically between OFC and dorsal medial pre-motor and cingulate cortex were attenuated in both patient groups. However, patients with OCD demonstrated more severe and extensive reductions of functional connectivity compared to SDIs. OCD and stimulant dependence are not identical at the level of brain functional systems but they have some important abnormalities in common compared with healthy volunteers. Orbitofrontal connectivity may serve as a human brain systems biomarker for compulsivity across diagnostic categories.