The Influence of Preprocessing Steps on Graph Theory Measures Derived from Resting State fMRI

The multiverse of data preprocessing and analysis in graph-based fMRI: A systematic literature review of analytical choices fed into a decision support tool for informed analysis

The large number of different analytical choices used by researchers is partly responsible for the challenge of replication in neuroimaging studies. For an exhaustive robustness analysis, knowledge of the full space of analytical options is essential. We conducted a systematic literature review to identify the analytical decisions in functional neuroimaging data preprocessing and analysis in the emerging field of cognitive network neuroscience. We found 61 different steps, with 17 of them having debatable parameter choices. Scrubbing, global signal regression, and spatial smoothing are among the controversial steps. There is no standardized order in which different steps are applied, and the parameter settings within several steps vary widely across studies. By aggregating the pipelines across studies, we propose three taxonomic levels to categorize analytical choices: 1) inclusion or exclusion of specific steps, 2) parameter tuning within steps, and 3) distinct sequencing of steps. We have developed a decision support application with high educational value called METEOR to facilitate access to the data in order to design well-informed robustness (multiverse) analysis.

Sex classification from functional brain connectivity: Generalization to multiple datasets

Machine learning (ML) approaches are increasingly being applied to neuroimaging data. Studies in neuroscience typically have to rely on a limited set of training data which may impair the generalizability of ML models. However, it is still unclear which kind of training sample is best suited to optimize generalization performance. In the present study, we systematically investigated the generalization performance of sex classification models trained on the parcelwise connectivity profile of either single samples or compound samples of two different sizes. Generalization performance was quantified in terms of mean across-sample classification accuracy and spatial consistency of accurately classifying parcels. Our results indicate that the generalization performance of parcelwise classifiers (pwCs) trained on single dataset samples is dependent on the specific test samples. Certain datasets seem to "match" in the sense that classifiers trained on a sample from one dataset achieved a high accuracy when tested on the respected other one and vice versa. The pwCs trained on the compound samples demonstrated overall highest generalization performance for all test samples, including one derived from a dataset not included in building the training samples. Thus, our results indicate that both a large sample size and a heterogeneous data composition of a training sample have a central role in achieving generalizable results.

Abnormal brain activity in lumbar disc herniation patients with chronic pain is associated with their clinical symptoms

Introduction

Lumbar disc herniation, a chronic degenerative disease, is one of the major contributors to chronic low back pain and disability. Although many studies have been conducted in the past on brain function in chronic low back pain, most of these studies did not classify chronic low back pain (cLBP) patients according to their etiology. The lack of etiologic classification may lead to inconsistencies between findings, and the correlation between differences in brain activation and clinical symptoms in patients with cLBP was less studied in the past.

Methods

In this study, 36 lumbar disc herniation patients with chronic low back pain (LDHCP) and 36 healthy controls (HCs) were included to study brain activity abnormalities in LDHCP. Visual analogue scale (VAS), oswestry disability index (ODI), self-rating anxiety scale (SAS), self-rating depression scale (SDS) were used to assess clinical symptoms.

Results

The results showed that LDHCP patients exhibited abnormally increased and diminished activation of brain regions compared to HCs. Correlation analysis showed that the amplitude of low frequency fluctuations (ALFF) in the left middle frontal gyrus is negatively correlated with SAS and VAS, while the right superior temporal gyrus is positively correlated with SAS and VAS, the dorsolateral left superior frontal gyrus and the right middle frontal gyrus are negatively correlated with VAS and SAS, respectively.

Conclusion

LDHCP patients have brain regions with abnormally increased and abnormally decreased activation compared to healthy controls. Furthermore, some of the abnormally activated brain regions were correlated with clinical pain or emotional symptoms.

Reorganization of the functional connectome from rest to a visual perception task in schizophrenia and bipolar disorder

Functional connectome organization is altered in schizophrenia (SZ) and bipolar disorder (BD). However, it remains unclear whether network reorganization during a task relative to rest is also altered in these disorders. This study examined connectome organization in patients with SZ (N = 43) and BD (N = 42) versus healthy controls (HC; N = 39) using fMRI data during a visual object-perception task and at rest. Graph analyses were conducted for the whole-brain network using indices selected a priori: three reflecting network segregation (clustering coefficient, local efficiency, modularity), two reflecting integration (characteristic path length, global efficiency). Group differences were limited to network segregation and were more evident in SZ (clustering coefficient, modularity) than in BD (clustering coefficient) compared to HC. State differences were found across groups for segregation (local efficiency) and integration (characteristic path length). There was no group-by-state interaction for any graph index. In summary, aberrant network organization compared to HC was confirmed, and was more evident in SZ than in BD. Yet, reorganization was largely intact in both disorders. These findings help to constrain models of dysconnection in SZ and BD, suggesting that the extent of functional dysconnectivity in these disorders tends to persist across changes in mental state.

Spontaneous brain activity, graph metrics, and head motion related to prospective post-traumatic stress disorder trauma-focused therapy response

Introduction

Trauma-focused psychotherapy for post-traumatic stress disorder (PTSD) is effective in about half of all patients. Investigating biological systems related to prospective treatment response is important to gain insight in mechanisms predisposing patients for successful intervention. We studied if spontaneous brain activity, brain network characteristics and head motion during the resting state are associated with future treatment success.

Methods

Functional magnetic resonance imaging scans were acquired from 46 veterans with PTSD around the start of treatment. Psychotherapy consisted of trauma-focused cognitive behavioral therapy (tf-CBT), eye movement desensitization and reprocessing (EMDR), or a combination thereof. After intervention, 24 patients were classified as treatment responders and 22 as treatment resistant. Differences between groups in spontaneous brain activity were evaluated using amplitude of low-frequency fluctuations (ALFF), while global and regional brain network characteristics were assessed using a minimum spanning tree (MST) approach. In addition, in-scanner head motion was assessed.

Results

No differences in spontaneous brain activity and global network characteristics were observed between the responder and non-responder group. The right inferior parietal lobule, right putamen and left superior parietal lobule had a more central position in the network in the responder group compared to the non-responder group, while the right dorsolateral prefrontal cortex (DLPFC), right inferior frontal gyrus and left inferior temporal gyrus had a less central position. In addition, responders showed less head motion.

Discussion

These results show that areas involved in executive functioning, attentional and action processes, learning, and visual-object processing, are related to prospective PTSD treatment response in veterans. In addition, these findings suggest that involuntary micromovements may be related to future treatment success.

Local efficiency analysis of restingstate functional brain network in methamphetamine users

This study set out to assess restingstate functional connectivity (rs-FN) and graph theorybased local efficiency within the left and right hemispheres of methamphetamine (MA) abusers. Functional brain networks of 19 MA abusers and 21 control participants were analyzed using restingstate fMRI. Graph edges in functional networks of the brain were defined and recurrence plot was used. We found that MA abuse may be accompanied by alterations of rs-FN within the defaultmode network (DMN), executive control network (ECN), and the salience network (SN) in both hemispheres of the brain. We also observed that such effects of MA may be correlated with duration of MA abuse and abstinence in many components of the DMN and SN. The results would seem to suggest that MAinduced alterations of local efficiency may, in part, account for maladaptive decision making, deficits in executive function and control over drug seeking/taking, and relapse.

Temporal Dynamics of Resting-state Functional Networks and Cognitive Functioning following Systemic Treatment for Breast Cancer

Many women with breast cancer suffer from a decline in memory and executive function, particularly after treatment with chemotherapy. Recent neuroimaging studies suggest that changes in network dynamics are fundamental in decline in these cognitive functions. This has, however, not yet been investigated in breast cancer patients. Using resting state functional magnetic resonance imaging, we prospectively investigated whether changes in dynamic functional connectivity were associated with changes in memory and executive function. We examined 34 breast cancer patients that received chemotherapy, 32 patients that did not receive chemotherapy, and 35 no-cancer controls. All participants were assessed prior to treatment and six months after completion of chemotherapy, or at similar intervals for the other groups. To assess memory and executive function, we used the Hopkins Verbal Learning Test – Immediate Recall and the Trail Making Test B, respectively. Using a sliding window approach, we then evaluated dynamic functional connectivity of resting state networks supporting memory and executive function, i.e. the default mode network and frontoparietal network, respectively. Next, we directly investigated the association between cognitive performance and dynamic functional connectivity. We found no group differences in cognitive performance or connectivity measures. The association between dynamic functional connectivity of the default mode network and memory differed significantly across groups. This was not the case for the frontoparietal network and executive function. This suggests that cancer and chemotherapy alter the role of dynamic functional connectivity in memory function. Further implications of these findings are discussed.

Future Directions for Chemosensory Connectomes: Best Practices and Specific Challenges

Ecological chemosensory stimuli almost always evoke responses in more than one sensory system. Moreover, any sensory processing takes place along a hierarchy of brain regions. So far, the field of chemosensory neuroimaging is dominated by studies that examine the role of brain regions in isolation. However, to completely understand neural processing of chemosensation, we must also examine interactions between regions. In general, the use of connectivity methods has increased in the neuroimaging field, providing important insights to physical sensory processing, such as vision, audition, and touch. A similar trend has been observed in chemosensory neuroimaging, however, these established techniques have largely not been rigorously applied to imaging studies on the chemical senses, leaving network insights overlooked. In this article, we first highlight some recent work in chemosensory connectomics and we summarize different connectomics techniques. Then, we outline specific challenges for chemosensory connectome neuroimaging studies. Finally, we review best practices from the general connectomics and neuroimaging fields. We recommend future studies to develop or use the following methods we perceive as key to improve chemosensory connectomics: (1) optimized study designs, (2) reporting guidelines, (3) consensus on brain parcellations, (4) consortium research, and (5) data sharing.

Functional connectome of brainstem nuclei involved in autonomic, limbic, pain and sensory processing in living humans from 7 Tesla resting state fMRI

Despite remarkable advances in mapping the functional connectivity of the cortex, the functional connectivity of subcortical regions is understudied in living humans. This is the case for brainstem nuclei that control vital processes, such as autonomic, limbic, nociceptive and sensory functions. This is because of the lack of precise brainstem nuclei localization, of adequate sensitivity and resolution in the deepest brain regions, as well as of optimized processing for the brainstem. To close the gap between the cortex and the brainstem, on 20 healthy subjects, we computed a correlation-based functional connectome of 15 brainstem nuclei involved in autonomic, limbic, nociceptive, and sensory function (superior and inferior colliculi, ventral tegmental area-parabrachial pigmented nucleus complex, microcellular tegmental nucleus-prabigeminal nucleus complex, lateral and medial parabrachial nuclei, vestibular and superior olivary complex, superior and inferior medullary reticular formation, viscerosensory motor nucleus, raphe magnus, pallidus, and obscurus, and parvicellular reticular nucleus - alpha part) with the rest of the brain. Specifically, we exploited 1.1mm isotropic resolution 7 Tesla resting-state fMRI, ad-hoc coregistration and physiological noise correction strategies, and a recently developed probabilistic template of brainstem nuclei. Further, we used 2.5mm isotropic resolution resting-state fMRI data acquired on a 3 Tesla scanner to assess the translatability of our results to conventional datasets. We report highly consistent correlation coefficients across subjects, confirming available literature on autonomic, limbic, nociceptive and sensory pathways, as well as high interconnectivity within the central autonomic network and the vestibular network. Interestingly, our results showed evidence of vestibulo-autonomic interactions in line with previous work. Comparison of 7 Tesla and 3 Tesla findings showed high translatability of results to conventional settings for brainstem-cortical connectivity and good yet weaker translatability for brainstem-brainstem connectivity. The brainstem functional connectome might bring new insight in the understanding of autonomic, limbic, nociceptive and sensory function in health and disease.

Functional ultrasound imaging: A useful tool for functional connectomics?

Functional ultrasound (fUS) is a hemodynamic-based functional neuroimaging technique, primarily used in animal models, that combines a high spatiotemporal resolution, a large field of view, and compatibility with behavior. These assets make fUS especially suited to interrogating brain activity at the systems level. In this review, we describe the technical capabilities offered by fUS and discuss how this technique can contribute to the field of functional connectomics. First, fUS can be used to study intrinsic functional connectivity, namely patterns of correlated activity between brain regions. In this area, fUS has made the most impact by following connectivity changes in disease models, across behavioral states, or dynamically. Second, fUS can also be used to map brain-wide pathways associated with an external event. For example, fUS has helped obtain finer descriptions of several sensory systems, and uncover new pathways implicated in specific behaviors. Additionally, combining fUS with direct circuit manipulations such as optogenetics is an attractive way to map the brain-wide connections of defined neuronal populations. Finally, technological improvements and the application of new analytical tools promise to boost fUS capabilities. As brain coverage and the range of behavioral contexts that can be addressed with fUS keep on increasing, we believe that fUS-guided connectomics will only expand in the future. In this regard, we consider the incorporation of fUS into multimodal studies combining diverse techniques and behavioral tasks to be the most promising research avenue.

Mapping Domain- and Age-Specific Functional Brain Activity for Children's Cognitive and Affective Development

The human brain undergoes rapid development during childhood, with significant improvement in a wide spectrum of cognitive and affective functions. Mapping domain- and age-specific brain activity patterns has important implications for characterizing the development of children's cognitive and affective functions. The current mainstay of brain templates is primarily derived from structural magnetic resonance imaging (MRI), and thus is not ideal for mapping children's cognitive and affective brain development. By integrating task-dependent functional MRI data from a large sample of 250 children (aged 7 to 12) across multiple domains and the latest easy-to-use and transparent preprocessing workflow, we here created a set of age-specific brain functional activity maps across four domains: attention, executive function, emotion, and risky decision-making. Moreover, we developed a toolbox named Developmental Brain Functional Activity maps across multiple domains that enables researchers to visualize and download domain- and age-specific brain activity maps for various needs. This toolbox and maps have been released on the Neuroimaging Informatics Tools and Resources Clearinghouse website ( http://www.nitrc.org/projects/dbfa ). Our study provides domain- and age-specific brain activity maps for future developmental neuroimaging studies in both healthy and clinical populations.

Resting-state networks in the course of aging-differential insights from studies across the lifespan vs. amongst the old

Resting-state functional connectivity (RSFC) has widely been used to examine reorganization of functional brain networks during normal aging. The extraction of generalizable age trends, however, is hampered by differences in methodological approaches, study designs and sample characteristics. Distinct age ranges of study samples thereby represent an important aspect between studies especially due to the increase in inter-individual variability over the lifespan. The current review focuses on comparing age-related differences in RSFC in the course of the whole adult lifespan versus later decades of life. We summarize and compare studies assessing age-related differences in within- and between-network RSFC of major resting-state brain networks. Differential effects of the factor age on resting-state networks can be identified when comparing studies focusing on younger versus older adults with studies investigating effects within the older adult population. These differential effects pertain to higher order and primary processing resting-state networks to a varying extent. Especially during later decades of life, other factors beyond age might come into play to understand the high inter-individual variability in RSFC.

Influence of structural and functional brain connectivity on age-related differences in fluid cognition

We used graph theoretical measures to investigate the hypothesis that structural brain connectivity constrains the influence of functional connectivity on the relation between age and fluid cognition. Across 143 healthy, community-dwelling adults 19-79 years of age, we estimated structural network properties from diffusion-weighted imaging and functional network properties from resting-state functional magnetic resonance imaging. We confirmed previous reports of age-related decline in the strength and efficiency of structural networks, as well as in the connectivity strength within and between structural network modules. Functional networks, in contrast, exhibited age-related decline only in system segregation, a measure of the distinctiveness among network modules. Aging was associated with decline in a composite measure of fluid cognition, particularly tests of executive function. Functional system segregation was a significant mediator of age-related decline in executive function. Structural network properties did not directly influence the age-related decline in functional system segregation. The raw correlational data underlying the graph theoretical measures indicated that structural connectivity exerts a limited constraint on age-related decline in functional connectivity.

The Modular Organization of Pain Brain Networks: An fMRI Graph Analysis Informed by Intracranial EEG

Intracranial EEG (iEEG) studies have suggested that the conscious perception of pain builds up from successive contributions of brain networks in less than 1 s. However, the functional organization of cortico-subcortical connections at the multisecond time scale, and its accordance with iEEG models, remains unknown. Here, we used graph theory with modular analysis of fMRI data from 60 healthy participants experiencing noxious heat stimuli, of whom 36 also received audio stimulation. Brain connectivity during pain was organized in four modules matching those identified through iEEG, namely: 1) sensorimotor (SM), 2) medial fronto-cingulo-parietal (default mode-like), 3) posterior parietal-latero-frontal (central executive-like), and 4) amygdalo-hippocampal (limbic). Intrinsic overlaps existed between the pain and audio conditions in high-order areas, but also pain-specific higher small-worldness and connectivity within the sensorimotor module. Neocortical modules were interrelated via “connector hubs” in dorsolateral frontal, posterior parietal, and anterior insular cortices, the antero-insular connector being most predominant during pain. These findings provide a mechanistic picture of the brain networks architecture and support fractal-like similarities between the micro-and macrotemporal dynamics associated with pain. The anterior insula appears to play an essential role in information integration, possibly by determining priorities for the processing of information and subsequent entrance into other points of the brain connectome.

Motion-Dependent Effects of Functional Magnetic Resonance Imaging Preprocessing Methodology on Global Functional Connectivity

Background: While functional magnetic resonance imaging (fMRI) has become an established noninvasive tool for studying brain activity in both healthy and diseased states, no broad consensus has been reached regarding preprocessing methodology. Furthermore, the relationship between variations in preprocessing and functional connectivity (FC) networks remains incompletely understood. Purpose: The aim of this study was to relate FC to (1) choices in preprocessing methodology and (2) subject motion. Methods: Clinical and MRI data were analyzed from healthy subjects acquired as part of the Autism Brain Imaging Data Exchange (ABIDE). Data were obtained from 508 healthy subjects. Data from subjects in the highest and lowest quartiles for motion were used to calculate the interaction between motion and preprocessing. Data were analyzed across four domains of fMRI preprocessing: (1) pipeline, (2) global signal regression (GSR), (3) bandpass filtering, and (4) anatomic atlas. For the FC network calculated from each preprocessing scheme, overall FC using Pearson correlation, as well as leaf fraction and diameter, was calculated for each subject, and statistical comparison was made across schemes using generalized estimating equations. Results: FC and global network properties were significantly affected by each preprocessing step, and each preprocessing step significantly interacted with subject motion to differentially affect global functional network properties, with GSR having the strongest effect. Conclusion: Preprocessing choices in fMRI studies influence overall FC and global network properties and can have motion-dependent effects. Impact statement Different parameter choices across the four domains of functional magnetic resonance imaging (fMRI) preprocessing analyzed in this study (global signal regression, bandpass filtering, anatomic atlas, and pipeline) were associated with differences in functional connectivity (FC) as well as global network properties in healthy subjects. Some fMRI preprocessing parameter choices resulted in motion-dependent effects on FC and global network properties.

Acquisition Duration in Resting-State Arterial Spin Labeling. How Long Is Enough?

Resting-state Arterial Spin Labeling (rs-ASL) is a rather confidential method compared to resting-state BOLD. As ASL allows to quantify the cerebral blood flow, unlike BOLD, rs-ASL can lead to significant clinical subject-scaled applications. Despite directly impacting clinical practicability and functional networks estimation, there is no standard for rs-ASL regarding the acquisition duration. Our work here focuses on assessing the feasibility of ASL as an rs-fMRI method and on studying the effect of the acquisition duration on the estimation of functional networks. To this end, we acquired a long 24 min 30 s rs-ASL sequence and investigated how estimations of six typical functional brain networks evolved with respect to the acquisition duration. Our results show that, after a certain acquisition duration, the estimations of all functional networks reach their best and are stabilized. Since, for clinical application, the acquisition duration should be the shortest possible, we suggest an acquisition duration of 14 min, i.e., 240 volumes with our sequence parameters, as it covers the functional networks estimation stabilization.

Effects of spatial smoothing on group-level differences in functional brain networks

Brain connectivity with functional magnetic resonance imaging (fMRI) is a popular approach for detecting differences between healthy and clinical populations. Before creating a functional brain network, the fMRI time series must undergo several preprocessing steps to control for artifacts and to improve data quality. However, preprocessing may affect the results in an undesirable way. Spatial smoothing, for example, is known to alter functional network structure. Yet, its effects on group-level network differences remain unknown. Here, we investigate the effects of spatial smoothing on the difference between patients and controls for two clinical conditions: autism spectrum disorder and bipolar disorder, considering fMRI data smoothed with Gaussian kernels (0–32 mm). We find that smoothing affects network differences between groups. For weighted networks, incrementing the smoothing kernel makes networks more different. For thresholded networks, larger smoothing kernels lead to more similar networks, although this depends on the network density. Smoothing also alters the effect sizes of the individual link differences. This is independent of the region of interest (ROI) size, but varies with link length. The effects of spatial smoothing are diverse, nontrivial, and difficult to predict. This has important consequences: The choice of smoothing kernel affects the observed network differences.

Spatial smoothing is a preprocessing tool commonly applied to reduce the amount of noise in functional magnetic resonance imaging (fMRI) data. However, smoothing is known to affect the outcomes of functional brain network analysis at the level of individual subjects in undesired ways. Here, we investigate how spatial smoothing affects the observed differences in brain network structure between subject groups. Using fMRI data from two clinical populations and healthy controls, we show that the between-group differences in network structure depend on the amount of spatial smoothing applied during preprocessing in a nontrivial way. The optimal level of spatial smoothing is difficult to define and probably depends on a set of analysis parameters. Therefore, we recommend applying spatial smoothing only after careful consideration.

Prefrontal-amygdala connectivity in trait anxiety and generalized anxiety disorder: Testing the boundaries between healthy and pathological worries

BACKGROUND

Current brain-based theoretical models of generalized anxiety disorder (GAD) suggest a dysfunction of amygdala-ventromedial prefrontal cortex emotional regulatory mechanisms. These alterations might be reflected by an altered resting state functional connectivity between both areas and could extend to vulnerable non-clinical samples such as high worriers without a GAD diagnosis. However, there is a lack of information in this regard.

METHODS

We investigated differences in resting state functional connectivity between the basolateral amygdala and the ventromedial prefrontal cortex (amygdala-vmPFC) in 28 unmedicated participants with GAD, 28 high-worriers and 28 low-worriers. We additionally explored selected clinical variables as predictors of amygdala-vmPFC connectivity, including anxiety sensitivity.

RESULTS

GAD participants presented higher left amygdala-vmPFC connectivity compared to both groups of non-GAD participants, and there were no differences between the latter two groups. In our exploratory analyses, concerns about the cognitive consequences of anxiety (the cognitive dimension of anxiety sensitivity) were found to be a significant predictor of the left amygdala-vmPFC connectivity.

LIMITATIONS

The cross-sectional nature of our study preclude us from assessing if functional connectivity measures and anxiety sensitivity scores entail an increased risk of GAD.

CONCLUSIONS

These results suggest a neurobiological qualitative distinction at the level of the amygdala-vmPFC emotional-regulatory system in GAD compared to non-GAD participants, either high- or low-worriers. At this neural level, they question previous hypotheses of continuity between high worries and GAD development. Instead, other anxiety traits such as anxiety sensitivity might confer a greater proneness to the amygdala-vmPFC connectivity alterations observed in GAD.

Resting State fMRI: Going Through the Motions

Resting state functional magnetic resonance imaging (rs-fMRI) has become an indispensable tool in neuroscience research. Despite this, rs-fMRI signals are easily contaminated by artifacts arising from movement of the head during data collection. The artifacts can be problematic even for motions on the millimeter scale, with complex spatiotemporal properties that can lead to substantial errors in functional connectivity estimates. Effective correction methods must be employed, therefore, to distinguish true functional networks from motion-related noise. Research over the last three decades has produced numerous correction methods, many of which must be applied in combination to achieve satisfactory data quality. Subject instruction, training, and mild restraints are helpful at the outset, but usually insufficient. Improvements come from applying multiple motion correction algorithms retrospectively after rs-fMRI data are collected, although residual artifacts can still remain in cases of elevated motion, which are especially prevalent in patient populations. Although not commonly adopted at present, “real-time” correction methods are emerging that can be combined with retrospective methods and that promise better correction and increased rs-fMRI signal sensitivity. While the search for the ideal motion correction protocol continues, rs-fMRI research will benefit from good disclosure practices, such as: (1) reporting motion-related quality control metrics to provide better comparison between studies; and (2) including motion covariates in group-level analyses to limit the extent of motion-related confounds when studying group differences.

Application of Graph Theory for Identifying Connectivity Patterns in Human Brain Networks: A Systematic Review

Background: Analysis of the human connectome using functional magnetic resonance imaging (fMRI) started in the mid-1990s and attracted increasing attention in attempts to discover the neural underpinnings of human cognition and neurological disorders. In general, brain connectivity patterns from fMRI data are classified as statistical dependencies (functional connectivity) or causal interactions (effective connectivity) among various neural units. Computational methods, especially graph theory-based methods, have recently played a significant role in understanding brain connectivity architecture. Objectives: Thanks to the emergence of graph theoretical analysis, the main purpose of the current paper is to systematically review how brain properties can emerge through the interactions of distinct neuronal units in various cognitive and neurological applications using fMRI. Moreover, this article provides an overview of the existing functional and effective connectivity methods used to construct the brain network, along with their advantages and pitfalls. Methods: In this systematic review, the databases Science Direct, Scopus, arXiv, Google Scholar, IEEE Xplore, PsycINFO, PubMed, and SpringerLink are employed for exploring the evolution of computational methods in human brain connectivity from 1990 to the present, focusing on graph theory. The Cochrane Collaboration's tool was used to assess the risk of bias in individual studies. Results: Our results show that graph theory and its implications in cognitive neuroscience have attracted the attention of researchers since 2009 (as the Human Connectome Project launched), because of their prominent capability in characterizing the behavior of complex brain systems. Although graph theoretical approach can be generally applied to either functional or effective connectivity patterns during rest or task performance, to date, most articles have focused on the resting-state functional connectivity. Conclusions: This review provides an insight into how to utilize graph theoretical measures to make neurobiological inferences regarding the mechanisms underlying human cognition and behavior as well as different brain disorders.

Kernel machine tests of association between brain networks and phenotypes

Applications of quantitative network analysis to functional brain connectivity have become popular in the last decade due to their ability to describe the general topological principles of brain networks. However, many issues arise when standard statistical analysis techniques are applied to functional magnetic resonance imaging (fMRI) connectivity maps. Frequently, summary measures of these maps, such as global efficiency and clustering coefficients, collapse the changing structures of graph topology from many scales to one. This can result in a loss of whole-brain spatio-temporal pattern information that may be significant in association and prediction analyses. Drawing from the electrical engineering field, the resistance perturbation distance is a quantification of similarity between graphs on the same vertex set that has been shown to identify changes in dynamic graphs, such as those from fMRI, while not being computationally expensive or result in a loss of information. This work proposes a novel kernel-based regression scheme that incorporates the resistance perturbation distance to better understand the association with biological phenotypes from fMRI using both simulated and real datasets.

Evaluation of different cerebrospinal fluid and white matter fMRI filtering strategies-Quantifying noise removal and neural signal preservation

This study examines the impact of using different cerebrospinal fluid (CSF) and white matter (WM) nuisance signals for data-driven filtering of functional magnetic resonance imaging (fMRI) data as a cleanup method before analyzing intrinsic brain fluctuations. The routinely used temporal signal-to-noise ratio metric is inappropriate for assessing fMRI filtering suitability, as it evaluates only the reduction of data variability and does not assess the preservation of signals of interest. We defined a new metric that evaluates the preservation of selected neural signal correlates, and we compared its performance with a recently published signal-noise separation metric. These two methods provided converging evidence of the unfavorable impact of commonly used filtering approaches that exploit higher numbers of principal components from CSF and WM compartments (typically 5 + 5 for CSF and WM, respectively). When using only the principal components as nuisance signals, using a lower number of signals results in a better performance (i.e., 1 + 1 performed best). However, there was evidence that this routinely used approach consisting of 1 + 1 principal components may not be optimal for filtering resting-state (RS) fMRI data, especially when RETROICOR filtering is applied during the data preprocessing. The evaluation of task data indicated the appropriateness of 1 + 1 principal components, but when RETROICOR was applied, there was a change in the optimal filtering strategy. The suggested change for extracting WM (and also CSF in RETROICOR-corrected RS data) is using local signals instead of extracting signals from a large mask using principal component analysis.

Application of Graph Theory for Identifying Connectivity Patterns in Human Brain Networks: A Systematic Review

Background: Analysis of the human connectome using functional magnetic resonance imaging (fMRI) started in the mid-1990s and attracted increasing attention in attempts to discover the neural underpinnings of human cognition and neurological disorders. In general, brain connectivity patterns from fMRI data are classified as statistical dependencies (functional connectivity) or causal interactions (effective connectivity) among various neural units. Computational methods, especially graph theory-based methods, have recently played a significant role in understanding brain connectivity architecture. Objectives: Thanks to the emergence of graph theoretical analysis, the main purpose of the current paper is to systematically review how brain properties can emerge through the interactions of distinct neuronal units in various cognitive and neurological applications using fMRI. Moreover, this article provides an overview of the existing functional and effective connectivity methods used to construct the brain network, along with their advantages and pitfalls. Methods: In this systematic review, the databases Science Direct, Scopus, arXiv, Google Scholar, IEEE Xplore, PsycINFO, PubMed, and SpringerLink are employed for exploring the evolution of computational methods in human brain connectivity from 1990 to the present, focusing on graph theory. The Cochrane Collaboration's tool was used to assess the risk of bias in individual studies. Results: Our results show that graph theory and its implications in cognitive neuroscience have attracted the attention of researchers since 2009 (as the Human Connectome Project launched), because of their prominent capability in characterizing the behavior of complex brain systems. Although graph theoretical approach can be generally applied to either functional or effective connectivity patterns during rest or task performance, to date, most articles have focused on the resting-state functional connectivity. Conclusions: This review provides an insight into how to utilize graph theoretical measures to make neurobiological inferences regarding the mechanisms underlying human cognition and behavior as well as different brain disorders.

GraphVar 2.0: A user-friendly toolbox for machine learning on functional connectivity measures

BACKGROUND

We previously presented GraphVar as a user-friendly MATLAB toolbox for comprehensive graph analyses of functional brain connectivity. Here we introduce a comprehensive extension of the toolbox allowing users to seamlessly explore easily customizable decoding models across functional connectivity measures as well as additional features.

NEW METHOD

GraphVar 2.0 provides machine learning (ML) model construction, validation and exploration. Machine learning can be performed across any combination of graph measures and additional variables, allowing for a flexibility in neuroimaging applications.

RESULTS

In addition to previously integrated functionalities, such as network construction and graph-theoretical analyses of brain connectivity with a high-speed general linear model (GLM), users can now perform customizable ML across connectivity matrices, graph measures and additionally imported variables. The new extension also provides parametric and nonparametric testing of classifier and regressor performance, data export, figure generation and high quality export.

COMPARISON WITH EXISTING METHODS

Compared to other existing toolboxes, GraphVar 2.0 offers (1) comprehensive customization, (2) an all-in-one user friendly interface, (3) customizable model design and manual hyperparameter entry, (4) interactive results exploration and data export, (5) automated queue system for modelling multiple outcome variables within the same session, (6) an easy to follow introductory review.

CONCLUSIONS

GraphVar 2.0 allows comprehensive, user-friendly exploration of encoding (GLM) and decoding (ML) modelling approaches on functional connectivity measures making big data neuroscience readily accessible to a broader audience of neuroimaging investigators.