Statistical Analysis of fMRI Time-Series: A Critical Review of the GLM Approach

Review of weighted exponential random graph models frameworks applied to neuroimaging

Neuro-imaging data can often be represented as statistical networks, especially for functional magnetic resonance imaging (fMRI) data, where brain regions are defined as nodes and the functional interactions between those regions are taken as edges. Such networks are commonly divided into classes depending on the type of edges, namely binary or weighted. A binary network means edges can either be present or absent. Whereas the edges of a weighted network are associated with weight values, and fMRI networks belong to weighted networks. Statistical methods are often adopted to analyse such networks, among which, the exponential random graph model (ERGM) is an important network analysis approach. Typically ERGMs are applied to binary networks, and weighted networks often need to be binarised by arbitrarily selecting a threshold value to define the presence of the edges, which can lead to non-robustness and loss of valuable edge weight information representing the strength of fMRI interaction in fMRI networks. While it is therefore important to gain deeper insight in adopting ERGM on weighted networks, there only exists a few different ERGM frameworks for weighted networks; some of these are not directly implementable on fMRI networks based on their original proposal. We systematically review, implement, analyse and compare five such frameworks via a simulation study and provide guidelines on each modelling framework as well as conclude the suitability of them on fMRI networks based on a range of criteria. We concluded that Multi-Layered ERGM is currently the most suitable framework.

Seeing more than the Tip of the Iceberg: Approaches to Subthreshold Effects in Functional Magnetic Resonance Imaging of the Brain

Many functional magnetic resonance imaging (fMRI) studies and presurgical mapping applications rely on mass-univariate inference with subsequent multiple comparison correction. Statistical results are frequently visualized as thresholded statistical maps. This approach has inherent limitations including the risk of drawing overly-selective conclusions based only on selective results passing such thresholds. This article gives an overview of both established and newly emerging scientific approaches to supplement such conventional analyses by incorporating information about subthreshold effects with the aim to improve interpretation of findings or leverage a wider array of information. Topics covered include neuroimaging data visualization, p-value histogram analysis and the related Higher Criticism approach for detecting rare and weak effects. Further examples from multivariate analyses and dedicated Bayesian approaches are provided.

Behavioral Bias for Exploration Is Associated with Enhanced Signaling in the Lateral and Medial Frontopolar Cortex

Should we keep doing what we know works for us, or should we risk trying something new as it could work even better? The exploration-exploitation dilemma is ubiquitous in daily life decision-making, and balancing between the two is crucial for adaptive behavior. Yet, we only have started to unravel the neurocognitive mechanisms that help us to find this balance in practice. Analyzing BOLD signals of healthy young adults during virtual foraging, we could show that a behavioral tendency for prolonged exploitation was associated with weakened signaling during exploration in central node points of the frontoparietal attention network, plus the frontopolar cortex. These results provide an important link between behavioral heuristics that we use to balance between exploitation and exploration and the brain function that supports shifts from one tendency to the other. Importantly, they stress that interindividual differences in behavioral strategies are reflected in differences in brain activity during exploration and should thus be more in the focus of basic research that aims at delineating general laws governing visual attention.

Explainability of three-dimensional convolutional neural networks for functional magnetic resonance imaging of Alzheimer's disease classification based on gradient-weighted class activation mapping

Currently, numerous studies focus on employing fMRI-based deep neural networks to diagnose neurological disorders such as Alzheimer's Disease (AD), yet only a handful have provided results regarding explainability. We address this gap by applying several prevalent explainability methods such as gradient-weighted class activation mapping (Grad-CAM) to an fMRI-based 3D-VGG16 network for AD diagnosis to improve the model's explainability. The aim is to explore the specific Region of Interest (ROI) of brain the model primarily focuses on when making predictions, as well as whether there are differences in these ROIs between AD and normal controls (NCs). First, we utilized multiple resting-state functional activity maps including ALFF, fALFF, ReHo, and VMHC to reduce the complexity of fMRI data, which differed from many studies that utilized raw fMRI data. Compared to methods utilizing raw fMRI data, this manual feature extraction approach may potentially alleviate the model's burden. Subsequently, 3D-VGG16 were employed for AD classification, where the final fully connected layers were replaced with a Global Average Pooling (GAP) layer, aimed at mitigating overfitting while preserving spatial information within the feature maps. The model achieved a maximum of 96.4% accuracy on the test set. Finally, several 3D CAM methods were employed to interpret the models. In the explainability results of the models with relatively high accuracy, the highlighted ROIs were primarily located in the precuneus and the hippocampus for AD subjects, while the models focused on the entire brain for NC. This supports current research on ROIs involved in AD. We believe that explaining deep learning models would not only provide support for existing research on brain disorders, but also offer important referential recommendations for the study of currently unknown etiologies.

Non-Cartesian 3D-SPARKLING vs Cartesian 3D-EPI encoding schemes for functional Magnetic Resonance Imaging at 7 Tesla

The quest for higher spatial and/or temporal resolution in functional MRI (fMRI) while preserving a sufficient temporal signal-to-noise ratio (tSNR) has generated a tremendous amount of methodological contributions in the last decade ranging from Cartesian vs. non-Cartesian readouts, 2D vs. 3D acquisition strategies, parallel imaging and/or compressed sensing (CS) accelerations and simultaneous multi-slice acquisitions to cite a few. In this paper, we investigate the use of a finely tuned version of 3D-SPARKLING. This is a non-Cartesian CS-based acquisition technique for high spatial resolution whole-brain fMRI. We compare it to state-of-the-art Cartesian 3D-EPI during both a retinotopic mapping paradigm and resting-state acquisitions at 1mm3 (isotropic spatial resolution). This study involves six healthy volunteers and both acquisition sequences were run on each individual in a randomly-balanced order across subjects. The performances of both acquisition techniques are compared to each other in regards to tSNR, sensitivity to the BOLD effect and spatial specificity. Our findings reveal that 3D-SPARKLING has a higher tSNR than 3D-EPI, an improved sensitivity to detect the BOLD contrast in the gray matter, and an improved spatial specificity. Compared to 3D-EPI, 3D-SPARKLING yields, on average, 7% more activated voxels in the gray matter relative to the total number of activated voxels.

A cautionary tale on the effects of different covariance structures in linear mixed effects modeling of fMRI data

With the steadily increasing abundance of longitudinal neuroimaging studies with large sample sizes and multiple repeated measures, questions arise regarding the appropriate modeling of variance and covariance. The current study examined the influence of standard classes of variance-covariance structures in linear mixed effects (LME) modeling of fMRI data from patients with pediatric mild traumatic brain injury (pmTBI; N = 181) and healthy controls (N = 162). During two visits, participants performed a cognitive control fMRI paradigm that compared congruent and incongruent stimuli. The hemodynamic response function was parsed into peak and late peak phases. Data were analyzed with a 4-way (GROUP×VISIT×CONGRUENCY×PHASE) LME using AFNI's 3dLME and compound symmetry (CS), autoregressive process of order 1 (AR1), and unstructured (UN) variance-covariance matrices. Voxel-wise results dramatically varied both within the cognitive control network (UN>CS for CONGRUENCY effect) and broader brain regions (CS>UN for GROUP:VISIT) depending on the variance-covariance matrix that was selected. Additional testing indicated that both model fit and estimated standard error were superior for the UN matrix, likely as a result of the modeling of individual terms. In summary, current findings suggest that the interpretation of results from complex designs is highly dependent on the selection of the variance-covariance structure using LME modeling.

Connectome caricatures: removing large-amplitude co-activation patterns in resting-state fMRI emphasizes individual differences

High-amplitude co-activation patterns are sparsely present during resting-state fMRI but drive functional connectivity 1-5 . Further, they resemble task activation patterns and are well-studied 3,5-10 . However, little research has characterized the remaining majority of the resting-state signal. In this work, we introduced caricaturing-a method to project resting-state data to a subspace orthogonal to a manifold of co-activation patterns estimated from the task fMRI data. Projecting to this subspace removes linear combinations of these co-activation patterns from the resting-state data to create Caricatured connectomes. We used rich task data from the Human Connectome Project (HCP) 11 and the UCLA Consortium for Neuropsychiatric Phenomics 12 to construct a manifold of task co-activation patterns. Caricatured connectomes were created by projecting resting-state data from the HCP and the Yale Test-Retest 13 datasets away from this manifold. Like caricatures, these connectomes emphasized individual differences by reducing between-individual similarity and increasing individual identification 14 . They also improved predictive modeling of brain-phenotype associations. As caricaturing removes group-relevant task variance, it is an initial attempt to remove task-like co-activations from rest. Therefore, our results suggest that there is a useful signal beyond the dominating co-activations that drive resting-state functional connectivity, which may better characterize the brain's intrinsic functional architecture.

Dynamic threat-reward neural processing under semi-naturalistic ecologically relevant scenarios

Studies of affective neuroscience have typically employed highly controlled, static experimental paradigms to investigate the neural underpinnings of threat and reward processing in the brain. Yet our knowledge of affective processing in more naturalistic settings remains limited. Specifically, affective studies generally examine threat and reward features separately and under brief time periods, despite the fact that in nature organisms are often exposed to the simultaneous presence of threat and reward features for extended periods. To study the neural mechanisms of threat and reward processing under distinct temporal profiles, we created a modified version of the PACMAN game that included these environmental features. We also conducted two automated meta-analyses to compare the findings from our semi-naturalistic paradigm to those from more constrained experiments. Overall, our results revealed a distributed system of regions sensitive to threat imminence and a less distributed system related to reward imminence, both of which exhibited overlap yet neither of which involved the amygdala. Additionally, these systems broadly overlapped with corresponding meta-analyses, with the notable absence of the amygdala in our findings. Together, these findings suggest a shared system for salience processing that reveals a heightened sensitivity toward environmental threats compared to rewards when both are simultaneously present in an environment. The broad correspondence of our findings to meta-analyses, consisting of more tightly controlled paradigms, illustrates how semi-naturalistic studies can corroborate previous findings in the literature while also potentially uncovering novel mechanisms resulting from the nuances and contexts that manifest in such dynamic environments.

Prediction of image interpretation cognitive ability under different mental workloads: a task-state fMRI study

Visual imaging experts play an important role in multiple fields, and studies have shown that the combination of functional magnetic resonance imaging and machine learning techniques can predict cognitive abilities, which provides a possible method for selecting individuals with excellent image interpretation skills. We recorded behavioral data and neural activity of 64 participants during image interpretation tasks under different workloads. Based on the comprehensive image interpretation ability, participants were divided into two groups. general linear model analysis showed that during image interpretation tasks, the high-ability group exhibited higher activation in middle frontal gyrus (MFG), fusiform gyrus, inferior occipital gyrus, superior parietal gyrus, inferior parietal gyrus, and insula compared to the low-ability group. The radial basis function Support Vector Machine (SVM) algorithm shows the most excellent performance in predicting participants' image interpretation abilities (Pearson correlation coefficient = 0.54, R2 = 0.31, MSE = 0.039, RMSE = 0.002). Variable importance analysis indicated that the activation features of the fusiform gyrus and MFG played an important role in predicting this ability. Our study revealed the neural basis related to image interpretation ability when exposed to different mental workloads. Additionally, our results demonstrated the efficacy of machine learning algorithms in extracting neural activation features to predict such ability.

Neuro-environmental interactions: a time sensitive matter

Introduction

The assessment of resting state (rs) neurophysiological dynamics relies on the control of sensory, perceptual, and behavioral environments to minimize variability and rule-out confounding sources of activation during testing conditions. Here, we investigated how temporally-distal environmental inputs, specifically metal exposures experienced up to several months prior to scanning, affect functional dynamics measured using rs functional magnetic resonance imaging (rs-fMRI).

Methods

We implemented an interpretable XGBoost-shapley additive explanation (SHAP) model that integrated information from multiple exposure biomarkers to predict rs dynamics in typically developing adolescents. In 124 participants (53% females, ages, 13-25 years) enrolled in the public health impact of metals exposure (PHIME) study, we measured concentrations of six metals (manganese, lead, chromium, copper, nickel, and zinc) in biological matrices (saliva, hair, fingernails, toenails, blood, and urine) and acquired rs-fMRI scans. Using graph theory metrics, we computed global efficiency (GE) in 111 brain areas (Harvard Oxford atlas). We used a predictive model based on ensemble gradient boosting to predict GE from metal biomarkers, adjusting for age and biological sex.

Results

Model performance was evaluated by comparing predicted versus measured GE. SHAP scores were used to evaluate feature importance. Measured versus predicted rs dynamics from our model utilizing chemical exposures as inputs were significantly correlated (p < 0.001, r = 0.36). Lead, chromium, and copper contributed most to the prediction of GE metrics.

Discussion

Our results indicate that a significant component of rs dynamics, comprising approximately 13% of observed variability in GE, is driven by recent metal exposures. These findings emphasize the need to estimate and control for the influence of past and current chemical exposures in the assessment and analysis of rs functional connectivity.

A novel technique for delineating the effect of variation in the learning rate on the neural correlates of reward prediction errors in model-based fMRI

Introduction

Computational models play an increasingly important role in describing variation in neural activation in human neuroimaging experiments, including evaluating individual differences in the context of psychiatric neuroimaging. In particular, reinforcement learning (RL) techniques have been widely adopted to examine neural responses to reward prediction errors and stimulus or action values, and how these might vary as a function of clinical status. However, there is a lack of consensus around the importance of the precision of free parameter estimation for these methods, particularly with regard to the learning rate. In the present study, I introduce a novel technique which may be used within a general linear model (GLM) to model the effect of mis-estimation of the learning rate on reward prediction error (RPE)-related neural responses.

Methods

Simulations employed a simple RL algorithm, which was used to generate hypothetical neural activations that would be expected to be observed in functional magnetic resonance imaging (fMRI) studies of RL. Similar RL models were incorporated within a GLM-based analysis method including derivatives, with individual differences in the resulting GLM-derived beta parameters being evaluated with respect to the free parameters of the RL model or being submitted to other validation analyses.

Results

Initial simulations demonstrated that the conventional approach to fitting RL models to RPE responses is more likely to reflect individual differences in a reinforcement efficacy construct (lambda) rather than learning rate (alpha). The proposed method, adding a derivative regressor to the GLM, provides a second regressor which reflects the learning rate. Validation analyses were performed including examining another comparable method which yielded highly similar results, and a demonstration of sensitivity of the method in presence of fMRI-like noise.

Conclusion

Overall, the findings underscore the importance of the lambda parameter for interpreting individual differences in RPE-coupled neural activity, and validate a novel neural metric of the modulation of such activity by individual differences in the learning rate. The method is expected to find application in understanding aberrant reinforcement learning across different psychiatric patient groups including major depression and substance use disorder.

Parametric information about eye movements is sent to the ears

Eye movements alter the relationship between the visual and auditory spatial scenes. Signals related to eye movements affect neural pathways from the ear through auditory cortex and beyond, but how these signals contribute to computing the locations of sounds with respect to the visual scene is poorly understood. Here, we evaluated the information contained in eye movement-related eardrum oscillations (EMREOs), pressure changes recorded in the ear canal that occur in conjunction with simultaneous eye movements. We show that EMREOs contain parametric information about horizontal and vertical eye displacement as well as initial/final eye position with respect to the head. The parametric information in the horizontal and vertical directions can be modeled as combining linearly, allowing accurate prediction of the EMREOs associated with oblique (diagonal) eye movements. Target location can also be inferred from the EMREO signals recorded during eye movements to those targets. We hypothesize that the (currently unknown) mechanism underlying EMREOs could impose a two-dimensional eye-movement-related transfer function on any incoming sound, permitting subsequent processing stages to compute the positions of sounds in relation to the visual scene.

A practical model-based segmentation approach for improved activation detection in single-subject functional magnetic resonance imaging studies

Functional magnetic resonance imaging (fMRI) maps cerebral activation in response to stimuli but this activation is often difficult to detect, especially in low-signal contexts and single-subject studies. Accurate activation detection can be guided by the fact that very few voxels are, in reality, truly activated and that these voxels are spatially localized, but it is challenging to incorporate both these facts. We address these twin challenges to single-subject and low-signal fMRI by developing a computationally feasible and methodologically sound model-based approach, implemented in the R package MixfMRI, that bounds the a priori expected proportion of activated voxels while also incorporating spatial context. An added benefit of our methodology is the ability to distinguish voxels and regions having different intensities of activation. Our suggested approach is evaluated in realistic two- and three-dimensional simulation experiments as well as on multiple real-world datasets. Finally, the value of our suggested approach in low-signal and single-subject fMRI studies is illustrated on a sports imagination experiment that is often used to detect awareness and improve treatment in patients in persistent vegetative state (PVS). Our ability to reliably distinguish activation in this experiment potentially opens the door to the adoption of fMRI as a clinical tool for the improved treatment and therapy of PVS survivors and other patients.

A generic framework for embedding human brain function with temporally correlated autoencoder

Learning an effective and compact representation of human brain function from high-dimensional fMRI data is crucial for studying the brain's functional organization. Traditional representation methods such as independent component analysis (ICA) and sparse dictionary learning (SDL) mainly rely on matrix decomposition which represents the brain function as spatial brain networks and the corresponding temporal patterns. The correspondence of those brain networks across individuals are built by viewing them as one-hot vectors and then performing the matching. However, those one-hot vectors do not encode the regularity and/or variability of different brains very well, and thus are limited in effectively representing the functional brain activities across individuals and among different time points. To address this problem, in this paper, we formulate the human brain functional representation as an embedding problem, and propose a novel embedding framework based on the Transformer model to encode the brain function in a compact, stereotyped and comparable latent space where the brain activities are represented as dense embedding vectors. We evaluate the proposed embedding framework on the publicly available Human Connectome Project (HCP) task fMRI dataset. The experiments on brain state prediction task indicate the effectiveness and generalizability of the learned embedding. We also explore the interpretability of the learned embedding from both spatial and temporal perspective. In general, our approach provides novel insights on representing the regularity and variability of human brain function in a general, comparable, and stereotyped latent space.

Increased interhemispheric functional connectivity during non-dominant hand movement in right-handed subjects

Hand preference is one of the behavioral expressions of lateralization in the brain. Previous fMRI studies showed the activation in several regions including the motor cortex and the cerebellum during single-hand movement. However, functional connectivity related to hand preference has not been investigated. Here, we used the generalized psychophysiological interaction (gPPI) approach to investigate the alteration of functional connectivity during single-hand movement from the resting state in right-hand subjects. The functional connectivity in interhemispheric motor-related regions including the supplementary motor area, the precentral gyrus, and the cerebellum was significantly increased during non-dominant hand movement, while functional connectivity was not increased during dominant hand movement. The general linear model (GLM) showed activation in contralateral supplementary motor area, contralateral precentral gyrus, and ipsilateral cerebellum during right- or left-hand movement. These results indicate that a combination of GLM and gPPI analysis can detect the lateralization of hand preference more clearly.

Covid-19 related cognitive, structural and functional brain changes among Italian adolescents and young adults: a multimodal longitudinal case-control study

Coronavirus disease 2019 (COVID-19) has been associated with brain functional, structural, and cognitive changes that persist months after infection. Most studies of the neurologic outcomes related to COVID-19 focus on severe infection and aging populations. Here, we investigated the neural activities underlying COVID-19 related outcomes in a case-control study of mildly infected youth enrolled in a longitudinal study in Lombardy, Italy, a global hotspot of COVID-19. All participants (13 cases, 27 controls, mean age 24 years) completed resting state functional (fMRI), structural MRI, cognitive assessments (CANTAB spatial working memory) at baseline (pre-COVID) and follow-up (post-COVID). Using graph theory eigenvector centrality (EC) and data-driven statistical methods, we examined differences in ECdelta (i.e., the difference in EC values pre- and post-COVID-19) and volumetricdelta (i.e., the difference in cortical volume of cortical and subcortical areas pre- and post-COVID) between COVID-19 cases and controls. We found that ECdeltasignificantly between COVID-19 and healthy participants in five brain regions; right intracalcarine cortex, right lingual gyrus, left hippocampus, left amygdala, left frontal orbital cortex. The left hippocampus showed a significant decrease in volumetricdelta between groups (p=0.041). The reduced ECdelta in the right amygdala associated with COVID-19 status mediated the association between COVID-19 and disrupted spatial working memory. Our results show persistent structural, functional and cognitive brain changes in key brain areas associated with olfaction and cognition. These results may guide treatment efforts to assess the longevity, reversibility and impact of the observed brain and cognitive changes following COVID-19.

Functional changes in neural mechanisms underlying post-traumatic stress disorder in World Trade Center responders

World Trade Center (WTC) responders exposed to traumatic and environmental stressors during rescue and recovery efforts have a high prevalence of chronic WTC-related post-traumatic stress disorder (WTC-PTSD). We investigated neural mechanisms underlying WTC-PTSD by applying eigenvector centrality (EC) metrics and data-driven methods on resting state functional magnetic resonance (fMRI). We identified how EC differences relate to WTC-exposure and behavioral symptoms. We found that connectivity differentiated significantly between WTC-PTSD and non-PTSD responders in nine brain regions, as these differences allowed an effective discrimination of PTSD and non-PTSD responders based solely on analysis of resting state data. Further, we found that WTC exposure duration (months on site) moderates the association between PTSD and EC values in two of the nine brain regions; the right anterior parahippocampal gyrus and the left amygdala (p = 0.010; p = 0.005, respectively, adjusted for multiple comparisons). Within WTC-PTSD, a dimensional measure of symptom severity was positively associated with EC values in the right anterior parahippocampal gyrus and brainstem. Functional neuroimaging can provide effective tools to identify neural correlates of diagnostic and dimensional indicators of PTSD.

Neuro-Environmental Interactions: a time sensitive matter

The assessment of resting state (rs) neurophysiological dynamics relies on the control of sensory, perceptual, and behavioral environments to minimize variability and rule-out confounding sources of activation during testing conditions. Here, we investigated how temporally-distal environmental inputs, specifically metal exposures experienced up to several months prior to scanning, affect functional dynamics measured using rs functional magnetic resonance imaging (rs-fMRI). We implemented an interpretable XGBoost-Shapley Additive Explanation (SHAP) model that integrated information from multiple exposure biomarkers to predict rs dynamics in typically developing adolescents. In 124 participants (53% females, ages: 13-25 years) enrolled in the Public Health Impact of Metals Exposure (PHIME) study, we measured concentrations of six metals (manganese, lead, chromium, cupper, nickel and zinc) in biological matrices (saliva, hair, fingernails, toenails, blood and urine) and acquired rs-fMRI scans. Using graph theory metrics, we computed global efficiency (GE) in 111 brain areas (Harvard Oxford Atlas). We used a predictive model based on ensemble gradient boosting to predict GE from metal biomarkers, adjusting for age and biological sex. Model performance was evaluated by comparing predicted versus measured GE. SHAP scores were used to evaluate feature importance. Measured versus predicted rs dynamics from our model utilizing chemical exposures as inputs were significantly correlated ( p < 0.001, r = 0.36). Lead, chromium, and copper contributed most to the prediction of GE metrics. Our results indicate that a significant component of rs dynamics, comprising approximately 13% of observed variability in GE, is driven by recent metal exposures. These findings emphasize the need to estimate and control for the influence of past and current chemical exposures in the assessment and analysis of rs functional connectivity.

MP-PCA denoising of fMRI time-series data can lead to artificial activation "spreading"

MP-PCA denoising has become the method of choice for denoising MRI data since it provides an objective threshold to separate the signal components from unwanted thermal noise components. In rodents, thermal noise in the coils is an important source of noise that can reduce the accuracy of activation mapping in fMRI. Further confounding this problem, vendor data often contains zero-filling and other post-processing steps that may violate MP-PCA assumptions. Here, we develop an approach to denoise vendor data and assess activation "spreading" caused by MP-PCA denoising in rodent task-based fMRI data. Data was obtained from N = 3 mice using conventional multislice and ultrafast acquisitions (1 s and 50 ms temporal resolution, respectively), during visual stimulation. MP-PCA denoising produced SNR gains of 64% and 39% and Fourier Spectral Amplitude (FSA) increases in BOLD maps of 9% and 7% for multislice and ultrafast data, respectively, when using a small [2 2] denoising window. Larger windows provided higher SNR and FSA gains with increased spatial extent of activation that may or may not represent real activation. Simulations showed that MP-PCA denoising can incur activation "spreading" with increased false positive rate and smoother functional maps due to local "bleeding" of principal components, and that the optimal denoising window for improved specificity of functional mapping, based on Dice score calculations, depends on the data's tSNR and functional CNR. This "spreading" effect applies also to another recently proposed low-rank denoising method (NORDIC), although to a lesser degree. Our results bode well for enhancing spatial and/or temporal resolution in future fMRI work, while taking into account the sensitivity/specificity trade-offs of low-rank denoising methods.

fNIRS is sensitive to leg activity in the primary motor cortex after systemic artifact correction

BACKGROUND

functional near-infrared spectroscopy (fNIRS) is an increasingly popular tool to study cortical activity during movement and gait that requires further validation. This study aimed to assess (1) whether fNIRS can detect the difficult-to-measure leg area of the primary motor cortex (M1) and distinguish it from the hand area; and (2) whether fNIRS can differentiate between automatic (i.e., not requiring one's attention) and non-automatic movement processes. Special attention was attributed to systemic artifacts (i.e., changes in blood pressure, heart rate, breathing) which were assessed and corrected by short channels, i.e., fNIRS channels which are mainly sensitive to superficial scalp hemodynamics.

METHODS

Twenty-three seated, healthy participants tapped four fingers on a keyboard or tapped the right foot on four squares on the floor in a specific order given by a 12-digit sequence (e.g., 434141243212). Two different sequences were executed: a beforehand learned (i.e., automatic) version and a newly learned (i.e., non-automatic) version. A 36-channel fNIRS device including 12 short channels covered multiple motor-related cortical areas including M1. The fNIRS data were analyzed with a general linear model (GLM). Correlation between the expected functional hemodynamic responses (i.e. task regressor) and the short channels (i.e. nuisance regressors), necessitated performing a separate short channel regression instead of integrating them in the GLM.

RESULTS

Consistent with the M1 somatotopy, we found significant HbO increases of very large effect size in the lateral M1 channels during finger tapping (Cohen's d = 1.35, p<0.001) and significant HbO increases of moderate effect size in the medial M1 channels during foot tapping (Cohen's d = 0.8, p<0.05). The cortical activity differences between automatic and non-automatic tasks were not significantly different. Importantly, leg movements produced large systemic fluctuations, which were adequately removed by the use of all available short channels.

DISCUSSION

Our results indicate that fNIRS is sensitive to leg activity in M1, though the sensitivity is lower than for finger activity and requires rigorous correction for systemic fluctuations. We furthermore highlight that systemic artifacts may result in an unreliable GLM analysis when short channels show signals that are similar to the expected hemodynamic responses.

Diverse Task Classification from Activation Patterns of Functional Neuro-Images Using Feature Fusion Module

One of the emerging fields in functional magnetic resonance imaging (fMRI) is the decoding of different stimulations. The underlying idea is to reveal the hidden representative signal patterns of various fMRI tasks for achieving high task-classification performance. Unfortunately, when multiple tasks are processed, performance remains limited due to several challenges, which are rarely addressed since the majority of the state-of-the-art studies cover a single neuronal activity task. Accordingly, the first contribution of this study is the collection and release of a rigorously acquired dataset, which contains cognitive, behavioral, and affective fMRI tasks together with resting state. After a comprehensive analysis of the pitfalls of existing systems on this new dataset, we propose an automatic multitask classification (MTC) strategy using a feature fusion module (FFM). FFM aims to create a unique signature for each task by combining deep features with time-frequency representations. We show that FFM creates a feature space that is superior for representing task characteristics compared to their individual use. Finally, for MTC, we test a diverse set of deep-models and analyze their complementarity. Our results reveal higher classification accuracy compared to benchmarks. Both the dataset and the code are accessible to researchers for further developments.

Explainable machine learning approach to predict and explain the relationship between task-based fMRI and individual differences in cognition

Despite decades of costly research, we still cannot accurately predict individual differences in cognition from task-based functional magnetic resonance imaging (fMRI). Moreover, aiming for methods with higher prediction is not sufficient. To understand brain-cognition relationships, we need to explain how these methods draw brain information to make the prediction. Here we applied an explainable machine-learning (ML) framework to predict cognition from task-based fMRI during the n-back working-memory task, using data from the Adolescent Brain Cognitive Development (n = 3,989). We compared 9 predictive algorithms in their ability to predict 12 cognitive abilities. We found better out-of-sample prediction from ML algorithms over the mass-univariate and ordinary least squares (OLS) multiple regression. Among ML algorithms, Elastic Net, a linear and additive algorithm, performed either similar to or better than nonlinear and interactive algorithms. We explained how these algorithms drew information, using SHapley Additive explanation, eNetXplorer, Accumulated Local Effects, and Friedman's H-statistic. These explainers demonstrated benefits of ML over the OLS multiple regression. For example, ML provided some consistency in variable importance with a previous study and consistency with the mass-univariate approach in the directionality of brain-cognition relationships at different regions. Accordingly, our explainable-ML framework predicted cognition from task-based fMRI with boosted prediction and explainability over standard methodologies.

Topological network properties of resting-state functional connectivity patterns are associated with metal mixture exposure in adolescents

Introduction

Adolescent exposure to neurotoxic metals adversely impacts cognitive, motor, and behavioral development. Few studies have addressed the underlying brain mechanisms of these metal-associated developmental outcomes. Furthermore, metal exposure occurs as a mixture, yet previous studies most often consider impacts of each metal individually. In this cross-sectional study, we investigated the relationship between exposure to neurotoxic metals and topological brain metrics in adolescents.

Methods

In 193 participants (53% females, ages: 15-25 years) enrolled in the Public Health Impact of Metals Exposure (PHIME) study, we measured concentrations of four metals (manganese, lead, copper, and chromium) in multiple biological media (blood, urine, hair, and saliva) and acquired resting-state functional magnetic resonance imaging scans. Using graph theory metrics, we computed global and local efficiency (global:GE; local:LE) in 111 brain areas (Harvard Oxford Atlas). We used weighted quantile sum (WQS) regression models to examine association between metal mixtures and each graph metric (GE or LE), adjusted for sex and age.

Results

We observed significant negative associations between the metal mixture and GE and LE [βGE = -0.076, 95% CI (-0.122, -0.031); βLE= -0.051, 95% CI (-0.095, -0.006)]. Lead and chromium measured in blood contributed most to this association for GE, while chromium measured in hair contributed the most for LE.

Discussion

Our results suggest that exposure to this metal mixture during adolescence reduces the efficiency of integrating information in brain networks at both local and global levels, informing potential neural mechanisms underlying the developmental toxicity of metals. Results further suggest these associations are due to combined joint effects to different metals, rather than to a single metal.

Sources of residual autocorrelation in multiband task fMRI and strategies for effective mitigation

Introduction

Analysis of task fMRI studies is typically based on using ordinary least squares within a voxel- or vertex-wise linear regression framework known as the general linear model. This use produces estimates and standard errors of the regression coefficients representing amplitudes of task-induced activations. To produce valid statistical inferences, several key statistical assumptions must be met, including that of independent residuals. Since task fMRI residuals often exhibit temporal autocorrelation, it is common practice to perform "prewhitening" to mitigate that dependence. Prewhitening involves estimating the residual correlation structure and then applying a filter to induce residual temporal independence. While theoretically straightforward, a major challenge in prewhitening for fMRI data is accurately estimating the residual autocorrelation at each voxel or vertex of the brain. Assuming a global model for autocorrelation, which is the default in several standard fMRI software tools, may under- or over-whiten in certain areas and produce differential false positive control across the brain. The increasing popularity of multiband acquisitions with faster temporal resolution increases the challenge of effective prewhitening because more complex models are required to accurately capture the strength and structure of autocorrelation. These issues are becoming more critical now because of a trend toward subject-level analysis and inference. In group-average or group-difference analyses, the within-subject residual correlation structure is accounted for implicitly, so inadequate prewhitening is of little real consequence. For individual subject inference, however, accurate prewhitening is crucial to avoid inflated or spatially variable false positive rates.

Methods

In this paper, we first thoroughly examine the patterns, sources and strength of residual autocorrelation in multiband task fMRI data. Second, we evaluate the ability of different autoregressive (AR) model-based prewhitening strategies to effectively mitigate autocorrelation and control false positives. We consider two main factors: the choice of AR model order and the level of spatial regularization of AR model coefficients, ranging from local smoothing to global averaging. We also consider determining the AR model order optimally at every vertex, but we do not observe an additional benefit of this over the use of higher-order AR models (e.g. (AR(6)). To overcome the computational challenge associated with spatially variable prewhitening, we developed a computationally efficient R implementation using parallelization and fast C++ backend code. This implementation is included in the open source R package BayesfMRI.

Results

We find that residual autocorrelation exhibits marked spatial variance across the cortex and is influenced by many factors including the task being performed, the specific acquisition protocol, mis-modeling of the hemodynamic response function, unmodeled noise due to subject head motion, and systematic individual differences. We also find that local regularization is much more effective than global averaging at mitigating autocorrelation. While increasing the AR model order is also helpful, it has a lesser effect than allowing AR coefficients to vary spatially. We find that prewhitening with an AR(6) model with local regularization is effective at reducing or even eliminating autocorrelation and controlling false positives.

Conclusion

Our analysis revealed dramatic spatial differences in autocorrelation across the cortex. This spatial topology is unique to each session, being influenced by the task being performed, the acquisition technique, various modeling choices, and individual differences. If not accounted for, these differences will result in differential false positive control and power across the cortex and across subjects.

Dual-task related frontal cerebral blood flow changes in older adults with mild cognitive impairment: A functional diffuse correlation spectroscopy study

Introduction

In a worldwide aging population with a high prevalence of motor and cognitive impairment, it is paramount to improve knowledge about underlying mechanisms of motor and cognitive function and their interplay in the aging processes.

Methods

We measured prefrontal cerebral blood flow (CBF) using functional diffuse correlation spectroscopy during motor and dual-task. We aimed to compare CBF changes among 49 older adults with and without mild cognitive impairment (MCI) during a dual-task paradigm (normal walk, 2- forward count walk, 3-backward count walk, obstacle negotiation, and heel tapping). Participants with MCI walked slower during the normal walk and obstacle negotiation compared to participants with normal cognition (NC), while gait speed during counting conditions was not different between the groups, therefore the dual-task cost was higher for participants with NC. We built a linear mixed effects model with CBF measures from the right and left prefrontal cortex.

Results

MCI (n = 34) showed a higher increase in CBF from the normal walk to the 2-forward count walk (estimate = 0.34, 95% CI [0.02, 0.66], p = 0.03) compared to participants with NC, related to a right- sided activation. Both groups showed a higher CBF during the 3-backward count walk compared to the normal walk, while only among MCI, CFB was higher during the 2-forward count walk.

Discussion

Our findings suggest a differential prefrontal hemodynamic pattern in older adults with MCI compared to their NC counterparts during the dual-task performance, possibly as a response to increasing attentional demand.

Multi-label classification of Alzheimer's disease stages from resting-state fMRI-based correlation connectivity data and deep learning

Alzheimer's disease is a neurodegenerative condition that gradually impairs cognitive abilities. Recently, various neuroimaging modalities and machine learning methods have surfaced to diagnose Alzheimer's disease. Resting-state fMRI is a neuroimaging modality that has been widely utilized to study brain activity related to neurodegenerative diseases. In literature, the previous studies are limited to the binary classification of Alzheimer's disease and Mild Cognitive Impairment. The application of computer-aided diagnosis for the numerous advancing phases of Alzheimer's disease, on the other hand, remains understudied. This research analyzes and presents methods for multi-label classification of six Alzheimer's stages using rs-fMRI and deep learning. The proposed model solves the multi-class classification problem by extracting the brain's functional connectivity networks from rs-fMRI data and employing two deep learning approaches, Stacked Sparse Autoencoder and Brain Connectivity Graph Convolutional Network. The suggested models' results were assessed using the k-fold cross-validation approach, and an average accuracy of 77.13% and 84.03% was reached for multi-label classification using Stacked Sparse Autoencoders and Brain Connectivity Based Convolutional Network, respectively. An analysis of brain regions was also performed by using the network's learned weights, leading to the conclusion that the precentral gyrus, frontal gyrus, lingual gyrus, and supplementary motor area are the significant brain regions of interest.

Improving the accuracy of single-trial fMRI response estimates using GLMsingle

Advances in artificial intelligence have inspired a paradigm shift in human neuroscience, yielding large-scale functional magnetic resonance imaging (fMRI) datasets that provide high-resolution brain responses to thousands of naturalistic visual stimuli. Because such experiments necessarily involve brief stimulus durations and few repetitions of each stimulus, achieving sufficient signal-to-noise ratio can be a major challenge. We address this challenge by introducing GLMsingle, a scalable, user-friendly toolbox available in MATLAB and Python that enables accurate estimation of single-trial fMRI responses (glmsingle.org). Requiring only fMRI time-series data and a design matrix as inputs, GLMsingle integrates three techniques for improving the accuracy of trial-wise general linear model (GLM) beta estimates. First, for each voxel, a custom hemodynamic response function (HRF) is identified from a library of candidate functions. Second, cross-validation is used to derive a set of noise regressors from voxels unrelated to the experiment. Third, to improve the stability of beta estimates for closely spaced trials, betas are regularized on a voxel-wise basis using ridge regression. Applying GLMsingle to the Natural Scenes Dataset and BOLD5000, we find that GLMsingle substantially improves the reliability of beta estimates across visually-responsive cortex in all subjects. Comparable improvements in reliability are also observed in a smaller-scale auditory dataset from the StudyForrest experiment. These improvements translate into tangible benefits for higher-level analyses relevant to systems and cognitive neuroscience. We demonstrate that GLMsingle: (i) helps decorrelate response estimates between trials nearby in time; (ii) enhances representational similarity between subjects within and across datasets; and (iii) boosts one-versus-many decoding of visual stimuli. GLMsingle is a publicly available tool that can significantly improve the quality of past, present, and future neuroimaging datasets sampling brain activity across many experimental conditions.

Fast Event-Related Mapping of Population Fingertip Tuning Properties in Human Sensorimotor Cortex at 7T

fMRI studies that investigate somatotopic tactile representations in the human cortex typically use either block or phase-encoded stimulation designs. Event-related (ER) designs allow for more flexible and unpredictable stimulation sequences than the other methods, but they are less efficient. Here, we compared an efficiency-optimized fast ER design (2.8-s average intertrial interval; ITI) to a conventional slow ER design (8-s average ITI) for mapping voxelwise fingertip tactile tuning properties in the sensorimotor cortex of six participants at 7 Tesla. The fast ER design yielded more reliable responses compared with the slow ER design, but with otherwise similar tuning properties. Concatenating the fast and slow ER data, we demonstrate in each individual brain the existence of two separate somatotopically-organized tactile representations of the fingertips, one in the primary somatosensory cortex (S1) on the postcentral gyrus, and the other shared across the motor and premotor cortices on the precentral gyrus. In both S1 and motor representations, fingertip selectivity decreased progressively, from narrowly-tuned Brodmann area (BA) 3b and BA4a, respectively, toward associative parietal and frontal regions that responded equally to all fingertips, suggesting increasing information integration along these two pathways. In addition, fingertip selectivity in S1 decreased from the cortical representation of the thumb to that of the pinky.

Capturing Brain-Cognition Relationship: Integrating Task-Based fMRI Across Tasks Markedly Boosts Prediction and Test-Retest Reliability

Capturing individual differences in cognition is central to human neuroscience. Yet our ability to estimate cognitive abilities via brain MRI is still poor in both prediction and reliability. Our study tested if this inability can be improved by integrating MRI signals across the whole brain and across modalities, including task-based functional MRI (tfMRI) of different tasks along with other non-task MRI modalities, such as structural MRI, resting-state functional connectivity. Using the Human Connectome Project (n=873, 473 females, after quality control), we directly compared predictive models comprising different sets of MRI modalities (e.g., seven tasks vs. non-task modalities). We applied two approaches to integrate multimodal MRI, stacked vs. flat models, and implemented 16 combinations of machine-learning algorithms. The stacked model integrating all modalities via stacking Elastic Net provided the best prediction (r=.57), relatively to other models tested, as well as excellent test-retest reliability (ICC=∼.85) in capturing general cognitive abilities. Importantly, compared to the stacked model integrating across non-task modalities (r=.27), the stacked model integrating tfMRI across tasks led to significantly higher prediction (r=.56) while still providing excellent test-retest reliability (ICC=∼.83). The stacked model integrating tfMRI across tasks was driven by frontal and parietal areas and by tasks that are cognition-related (working-memory, relational processing, and language). This result is consistent with the parieto-frontal integration theory of intelligence. Accordingly, our results contradict the recently popular notion that tfMRI is not reliable enough to capture individual differences in cognition. Instead, our study suggests that tfMRI, when used appropriately (i.e., by drawing information across the whole brain and across tasks and by integrating with other modalities), provides predictive and reliable sources of information for individual differences in cognitive abilities, more so than non-task modalities.

Longitudinal surface-based spatial Bayesian GLM reveals complex trajectories of motor neurodegeneration in ALS

Longitudinal fMRI studies hold great promise for the study of neurodegenerative diseases, development and aging, but realizing their full potential depends on extracting accurate fMRI-based measures of brain function and organization in individual subjects over time. This is especially true for studies of rare, heterogeneous and/or rapidly progressing neurodegenerative diseases. These often involve small samples with heterogeneous functional features, making traditional group-difference analyses of limited utility. One such disease is amyotrophic lateral sclerosis (ALS), a severe disease resulting in extreme loss of motor function and eventual death. Here, we use an advanced individualized task fMRI analysis approach to analyze a rich longitudinal dataset containing 190 hand clench fMRI scans from 16 ALS patients (78 scans) and 22 age-matched healthy controls (112 scans) Specifically, we adopt our cortical surface-based spatial Bayesian general linear model (GLM), which has high power and precision to detect activations in individual subjects, and propose a novel longitudinal extension to leverage information shared across visits. We perform all analyses in native surface space to preserve individua anatomical and functional features. Using mixed-effects models to subsequently study the relationship between size of activation and ALS disease progression, we observe for the first time an inverted U-shaped trajectory o motor activations: at relatively mild motor disability we observe enlarging activations, while at higher levels of motor disability we observe severely diminished activation, reflecting progression toward complete loss of motor function. We further observe distinct trajectories depending on clinical progression rate, with faster progressors exhibiting more extreme changes at an earlier stage of disability. These differential trajectories suggest that initial hyper-activation is likely attributable to loss of inhibitory neurons, rather than functional compensation as earlier assumed. These findings substantially advance scientific understanding of the ALS disease process. This study also provides the first real-world example of how surface-based spatial Bayesian analysis of task fMRI can further scientific understanding of neurodegenerative disease and other phenomena. The surface-based spatial Bayesian GLM is implemented in the BayesfMRI R package

Ultrasonic Deep Brain Neuromodulation in Acute Disorders of Consciousness: A Proof-of-Concept

The promotion of recovery in patients who have entered a disorder of consciousness (DOC; e.g., coma or vegetative states) following severe brain injury remains an enduring medical challenge despite an ever-growing scientific understanding of these conditions. Indeed, recent work has consistently implicated altered cortical modulation by deep brain structures (e.g., the thalamus and the basal ganglia) following brain damage in the arising of, and recovery from, DOCs. The (re)emergence of low-intensity focused ultrasound (LIFU) neuromodulation may provide a means to selectively modulate the activity of deep brain structures noninvasively for the study and treatment of DOCs. This technique is unique in its combination of relatively high spatial precision and noninvasive implementation. Given the consistent implication of the thalamus in DOCs and prior results inducing behavioral recovery through invasive thalamic stimulation, here we applied ultrasound to the central thalamus in 11 acute DOC patients, measured behavioral responsiveness before and after sonication, and applied functional MRI during sonication. With respect to behavioral responsiveness, we observed significant recovery in the week following thalamic LIFU compared with baseline. With respect to functional imaging, we found decreased BOLD signals in the frontal cortex and basal ganglia during LIFU compared with baseline. In addition, we also found a relationship between altered connectivity of the sonicated thalamus and the degree of recovery observed post-LIFU.

Brain Decoding Using fMRI Images for Multiple Subjects through Deep Learning

Substantial information related to human cerebral conditions can be decoded through various noninvasive evaluating techniques like fMRI. Exploration of the neuronal activity of the human brain can divulge the thoughts of a person like what the subject is perceiving, thinking, or visualizing. Furthermore, deep learning techniques can be used to decode the multifaceted patterns of the brain in response to external stimuli. Existing techniques are capable of exploring and classifying the thoughts of the human subject acquired by the fMRI imaging data. fMRI images are the volumetric imaging scans which are highly dimensional as well as require a lot of time for training when fed as an input in the deep learning network. However, the hassle for more efficient learning of highly dimensional high-level features in less training time and accurate interpretation of the brain voxels with less misclassification error is needed. In this research, we propose an improved CNN technique where features will be functionally aligned. The optimal features will be selected after dimensionality reduction. The highly dimensional feature vector will be transformed into low dimensional space for dimensionality reduction through autoadjusted weights and combination of best activation functions. Furthermore, we solve the problem of increased training time by using Swish activation function, making it denser and increasing efficiency of the model in less training time. Finally, the experimental results are evaluated and compared with other classifiers which demonstrated the supremacy of the proposed model in terms of accuracy.

A specific brain network for a social map in the human brain

Individuals use social information to guide social interactions and to update relationships along multiple social dimensions. However, it is unclear what neural basis underlies this process of abstract "social navigation". In the current study, we recruited twenty-nine participants who performed a choose-your-own-adventure game in which they interacted with fictional characters during fMRI scanning. Using a whole-brain GLM approach, we found that vectors encoding two-dimensional information about the relationships predicted BOLD responses in the hippocampus and the precuneus, replicating previous work. We also explored whether these geometric representations were related to key brain regions previously identified in physical and abstract spatial navigation studies, but we did not find involvement of the entorhinal cortex, parahippocampal gyrus or the retrosplenial cortex. Finally, we used psychophysiological interaction analysis and identified a network of regions that correlated during participants' decisions, including the left posterior hippocampus, precuneus, dorsolateral prefrontal cortex (dlPFC), and the insula. Our findings suggest a brain network for social navigation in multiple abstract, social dimensions that includes the hippocampus, precuneus, dlPFC, and insula.

Binocular Integrated Visual Field Deficits Are Associated With Changes in Local Network Function in Primary Open-Angle Glaucoma: A Resting-State fMRI Study

In glaucoma participants, both structural and functional brain changes have been observed, but we still have insufficient understanding of how these changes also affect the integrity of cortical functional networks, and how these changes relate to visual function. This is relevant, as functional network integrity may affect the applicability of future treatments, as well as the options for rehabilitation or training. Here, we compare global and local functional connectivity in local and global brain networks between glaucoma and control participants. Moreover, we study the relationship between functional connectivity and visual field (VF) loss. For our study, 20 subjects with primary open-angle glaucoma (POAG) and 24 age-similar healthy participants were recruited to undergo an ophthalmic assessment followed by two resting-state (RS) (f)MRI scans. For each scan and for each group, the ROIs with eigenvector centrality (EC) values higher than the 95th percentile were considered the most central brain regions (“hubs”). Hubs for which we found a significant difference in EC in both scans between glaucoma and healthy participants were considered to provide evidence for network changes. In addition, we tested the notion that a brain region's hub function in POAG might relate to the severity of a participant's VF defect, irrespective of which eye contributed mostly to this. To determine this, for each participant, eye-independent scores were derived for: (1) sensitivity of the worse eye – indicating disease severity, (2) sensitivity of both eyes combined – with one eye potentially compensating for loss in the other, or (3) difference in eye sensitivity – potentially requiring additional network interactions. By correlating each of these VF scores and the EC values, we assessed whether VF defects could be associated with centrality alterations in POAG. Our results show that no functional connectivity disruptions were found at the global brain level in POAG participants. This indicates that in glaucoma global brain network communication is preserved. Furthermore, for the Lingual Gyrus, identified as a brain hub, we found a positive correlation between the EC value and the VF sensitivity of both eyes combined. The fact that reduced local network functioning is associated with reduced binocular VF sensitivity suggests the presence of local brain reorganization that has a bearing on functional visual abilities.

A Computational Neural Model for Mapping Degenerate Neural Architectures

Degeneracy in biological systems refers to a many-to-one mapping between physical structures and their functional (including psychological) outcomes. Despite the ubiquity of the phenomenon, traditional analytical tools for modeling degeneracy in neuroscience are extremely limited. In this study, we generated synthetic datasets to describe three situations of degeneracy in fMRI data to demonstrate the limitations of the current univariate approach. We describe a novel computational approach for the analysis referred to as neural topographic factor analysis (NTFA). NTFA is designed to capture variations in neural activity across task conditions and participants. The advantage of this discovery-oriented approach is to reveal whether and how experimental trials and participants cluster into task conditions and participant groups. We applied NTFA on simulated data, revealing the appropriate degeneracy assumption in all three situations and demonstrating NTFA's utility in uncovering degeneracy. Lastly, we discussed the importance of testing degeneracy in fMRI data and the implications of applying NTFA to do so.

Structural and functional connectivity of the ascending arousal network for prediction of outcome in patients with acute disorders of consciousness

To determine the role of early acquisition of blood oxygen level-dependent (BOLD) signals and diffusion tensor imaging (DTI) for analysis of the connectivity of the ascending arousal network (AAN) in predicting neurological outcomes after acute traumatic brain injury (TBI), cardiopulmonary arrest (CPA), or stroke. A prospective analysis of 50 comatose patients was performed during their ICU stay. Image processing was conducted to assess structural and functional connectivity of the AAN. Outcomes were evaluated after 3 and 6 months. Nineteen patients (38%) had stroke, 18 (36%) CPA, and 13 (26%) TBI. Twenty-three patients were comatose (44%), 11 were in a minimally conscious state (20%), and 16 had unresponsive wakefulness syndrome (32%). Univariate analysis demonstrated that measurements of diffusivity, functional connectivity, and numbers of fibers in the gray matter, white matter, whole brain, midbrain reticular formation, and pontis oralis nucleus may serve as predictive biomarkers of outcome depending on the diagnosis. Multivariate analysis demonstrated a correlation of the predicted value and the real outcome for each separate diagnosis and for all the etiologies together. Findings suggest that the above imaging biomarkers may have a predictive role for the outcome of comatose patients after acute TBI, CPA, or stroke.

Statistical Learning of Frequent Distractor Locations in Visual Search Involves Regional Signal Suppression in Early Visual Cortex

Observers can learn locations where salient distractors appear frequently to reduce potential interference-an effect attributed to better suppression of distractors at frequent locations. But how distractor suppression is implemented in the visual cortex and within the frontoparietal attention networks remains unclear. We used fMRI and a regional distractor-location learning paradigm with two types of distractors defined in either the same (orientation) or a different (color) dimension to the target to investigate this issue. fMRI results showed that BOLD signals in early visual cortex were significantly reduced for distractors (as well as targets) occurring at the frequent versus rare locations, mirroring behavioral patterns. This reduction was more robust with same-dimension distractors. Crucially, behavioral interference was correlated with distractor-evoked visual activity only for same- (but not different-) dimension distractors. Moreover, with different- (but not same-) dimension distractors, a color-processing area within the fusiform gyrus was activated more when a distractor was present in the rare region versus being absent and more with a distractor in the rare versus frequent locations. These results support statistical learning of frequent distractor locations involving regional suppression in early visual cortex and point to differential neural mechanisms of distractor handling with different- versus same-dimension distractors.

Comparing Autoregressive and Network Features for Classification of Depression and Anxiety

Autocorrelation in functional MRI (fMRI) time series has been studied for decades, mostly considered as noise in the time series which is removed via prewhitening with an autoregressive model. Recent results suggest that the coefficients of an autoregressive model t to fMRI data may provide an indicator of underlying brain activity, suggesting that prewhitening could be removing important diagnostic information. This paper explores the explanatory value of these autoregressive features extracted from fMRI by considering the use of these features in a classification task. As a point of comparison, functional network based features are extracted from the same data and used in the same classification task. We find that in most cases, network based features provide better classification accuracy. However, using principal component analysis to combine network based features and autoregressive features for classification based on a support vector machine provides improved classification accuracy compared to single features or network features, suggesting that when properly combined there may be additional information to be gained from autoregressive features.

Systematic Review of Functional MRI Applications for Psychiatric Disease Subtyping

Background: Psychiatric disorders have been historically classified using symptom information alone. Recently, there has been a dramatic increase in research interest not only in identifying the mechanisms underlying defined pathologies but also in redefining their etiology. This is particularly relevant for the field of personalized medicine, which searches for data-driven approaches to improve diagnosis, prognosis, and treatment selection for individual patients. Methods: This review aims to provide a high-level overview of the rapidly growing field of functional magnetic resonance imaging (fMRI) from the perspective of unsupervised machine learning applications for disease subtyping. Following the PRISMA guidelines for protocol reproducibility, we searched the PubMed database for articles describing functional MRI applications used to obtain, interpret, or validate psychiatric disease subtypes. We also employed the active learning framework ASReview to prioritize publications in a machine learning-guided way. Results: From the 20 studies that met the inclusion criteria, five used functional MRI data to interpret symptom-derived disease clusters, four used it to interpret clusters derived from biomarker data other than fMRI itself, and 11 applied clustering techniques involving fMRI directly. Major depression disorder and schizophrenia were the two most frequently studied pathologies (35% and 30% of the retrieved studies, respectively), followed by ADHD (15%), psychosis as a whole (10%), autism disorder (5%), and the consequences of early exposure to violence (5%). Conclusions: The increased interest in personalized medicine and data-driven disease subtyping also extends to psychiatric disorders. However, to date, this subfield is at an incipient exploratory stage, and all retrieved studies were mostly proofs of principle where further validation and increased sample sizes are craved for. Whereas results for all explored diseases are inconsistent, we believe this reflects the need for concerted, multisite data collection efforts with a strong focus on measuring the generalizability of results. Finally, whereas functional MRI is the best way of measuring brain function available to date, its low signal-to-noise ratio and elevated monetary cost make it a poor clinical alternative. Even with technology progressing and costs decreasing, this might incentivize the search for more accessible, clinically ready functional proxies in the future.

Design of Deep Learning Model for Task-Evoked fMRI Data Classification

Machine learning methods have been successfully applied to neuroimaging signals, one of which is to decode specific task states from functional magnetic resonance imaging (fMRI) data. In this paper, we propose a model that simultaneously utilizes characteristics of both spatial and temporal sequential information of fMRI data with deep neural networks to classify the fMRI task states. We designed a convolution network module and a recurrent network module to extract the spatial and temporal features of fMRI data, respectively. In particular, we also add the attention mechanism to the recurrent network module, which more effectively highlights the brain activation state at the moment of reaction. We evaluated the model using task-evoked fMRI data from the Human Connectome Project (HCP) dataset, the classification accuracy got 94.31%, and the experimental results have shown that the model can effectively distinguish the brain states under different task stimuli.

A Motion Artifact Correction Procedure for fNIRS Signals Based on Wavelet Transform and Infrared Thermography Video Tracking

Functional near infrared spectroscopy (fNIRS) is a neuroimaging technique that allows to monitor the functional hemoglobin oscillations related to cortical activity. One of the main issues related to fNIRS applications is the motion artefact removal, since a corrupted physiological signal is not correctly indicative of the underlying biological process. A novel procedure for motion artifact correction for fNIRS signals based on wavelet transform and video tracking developed for infrared thermography (IRT) is presented. In detail, fNIRS and IRT were concurrently recorded and the optodes' movement was estimated employing a video tracking procedure developed for IRT recordings. The wavelet transform of the fNIRS signal and of the optodes' movement, together with their wavelet coherence, were computed. Then, the inverse wavelet transform was evaluated for the fNIRS signal excluding the frequency content corresponding to the optdes' movement and to the coherence in the epochs where they were higher with respect to an established threshold. The method was tested using simulated functional hemodynamic responses added to real resting-state fNIRS recordings corrupted by movement artifacts. The results demonstrated the effectiveness of the procedure in eliminating noise, producing results with higher signal to noise ratio with respect to another validated method.

The modulation of brain network integration and arousal during exploration

There is growing interest in how neuromodulators shape brain networks. Recent neuroimaging studies provide evidence that brainstem arousal systems, such as the locus coeruleus-norepinephrine system (LC-NE), influence functional connectivity and brain network topology, suggesting they have a role in flexibly reconfiguring brain networks in order to adapt behavior and cognition to environmental demands. To date, however, the relationship between brainstem arousal systems and functional connectivity has not been assessed within the context of a task with an established relationship between arousal and behavior, with most prior studies relying on incidental variations in arousal or pharmacological manipulation and static brain networks constructed over long periods of time. These factors have likely contributed to a heterogeneity of effects across studies. To address these issues, we took advantage of the association between LC-NE-linked arousal and exploration to probe the relationships between exploratory choice, arousal—as measured indirectly via pupil diameter—and brain network dynamics. Exploration in a bandit task was associated with a shift toward fewer, more weakly connected modules that were more segregated in terms of connectivity and topology but more integrated with respect to the diversity of cognitive systems represented in each module. Functional connectivity strength decreased, and changes in connectivity were correlated with changes in pupil diameter, in line with the hypothesis that brainstem arousal systems influence the dynamic reorganization of brain networks. More broadly, we argue that carefully aligning dynamic network analyses with task designs can increase the temporal resolution at which behaviorally- and cognitively-relevant modulations can be identified, and offer these results as a proof of concept of this approach.

Data Processing in Functional Near-Infrared Spectroscopy (fNIRS) Motor Control Research

FNIRS pre-processing and processing methodologies are very important-how a researcher chooses to process their data can change the outcome of an experiment. The purpose of this review is to provide a guide on fNIRS pre-processing and processing techniques pertinent to the field of human motor control research. One hundred and twenty-three articles were selected from the motor control field and were examined on the basis of their fNIRS pre-processing and processing methodologies. Information was gathered about the most frequently used techniques in the field, which included frequency cutoff filters, wavelet filters, smoothing filters, and the general linear model (GLM). We discuss the methodologies of and considerations for these frequently used techniques, as well as those for some alternative techniques. Additionally, general considerations for processing are discussed.

Evidence of Neurovascular Un-Coupling in Mild Alzheimer's Disease through Multimodal EEG-fNIRS and Multivariate Analysis of Resting-State Data

Alzheimer’s disease (AD) is associated with modifications in cerebral blood perfusion and autoregulation. Hence, neurovascular coupling (NC) alteration could become a biomarker of the disease. NC might be assessed in clinical settings through multimodal electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS). Multimodal EEG-fNIRS was recorded at rest in an ambulatory setting to assess NC and to evaluate the sensitivity and specificity of the methodology to AD. Global NC was evaluated with a general linear model (GLM) framework by regressing whole-head EEG power envelopes in three frequency bands (theta, alpha and beta) with average fNIRS oxy- and deoxy-hemoglobin concentration changes in the frontal and prefrontal cortices. NC was lower in AD compared to healthy controls (HC) with significant differences in the linkage of theta and alpha bands with oxy- and deoxy-hemoglobin, respectively (p = 0.028 and p = 0.020). Importantly, standalone EEG and fNIRS metrics did not highlight differences between AD and HC. Furthermore, a multivariate data-driven analysis of NC between the three frequency bands and the two hemoglobin species delivered a cross-validated classification performance of AD and HC with an Area Under the Curve, AUC = 0.905 (p = 2.17 × 10⁻⁵). The findings demonstrate that EEG-fNIRS may indeed represent a powerful ecological tool for clinical evaluation of NC and early identification of AD.

Underlying neurological mechanisms associated with symptomatic convergence insufficiency

Convergence insufficiency (CI) is the most common binocular vision problem, associated with blurred/double vision, headaches, and sore eyes that are exacerbated when doing prolonged near work, such as reading. The Convergence Insufficiency Neuro-mechanism Adult Population Study (NCT03593031) investigates the mechanistic neural differences between 50 binocularly normal controls (BNC) and 50 symptomatic CI participants by examining the fast and slow fusional disparity vergence systems. The fast fusional system is preprogrammed and is assessed with convergence peak velocity. The slow fusional system optimizes vergence effort and is assessed by measuring the phoria adaptation magnitude and rate. For the fast fusional system, significant differences are observed between the BNC and CI groups for convergence peak velocity, final position amplitude, and functional imaging activity within the secondary visual cortex, right cuneus, and oculomotor vermis. For the slow fusional system, the phoria adaptation magnitude and rate, and the medial cuneus functional activity, are significantly different between the groups. Significant correlations are observed between vergence peak velocity and right cuneus functional activity (p = 0.002) and the rate of phoria adaptation and medial cuneus functional activity (p = 0.02). These results map the brain-behavior of vergence. Future therapeutic interventions may consider implementing procedures that increase cuneus activity for this debilitating disorder.

Real time and delayed effects of subcortical low intensity focused ultrasound

Deep brain nuclei are integral components of large-scale circuits mediating important cognitive and sensorimotor functions. However, because they fall outside the domain of conventional non-invasive neuromodulatory techniques, their study has been primarily based on neuropsychological models, limiting the ability to fully characterize their role and to develop interventions in cases where they are damaged. To address this gap, we used the emerging technology of non-invasive low-intensity focused ultrasound (LIFU) to directly modulate left lateralized basal ganglia structures in healthy volunteers. During sonication, we observed local and distal decreases in blood oxygenation level dependent (BOLD) signal in the targeted left globus pallidus (GP) and in large-scale cortical networks. We also observed a generalized decrease in relative perfusion throughout the cerebrum following sonication. These results show, for the first time using functional MRI data, the ability to modulate deep-brain nuclei using LIFU while measuring its local and global consequences, opening the door for future applications of subcortical LIFU.

An analysis framework for the integration of broadband NIRS and EEG to assess neurovascular and neurometabolic coupling

With the rapid growth of optical-based neuroimaging to explore human brain functioning, our research group has been developing broadband Near Infrared Spectroscopy (bNIRS) instruments, a technological extension to functional Near Infrared Spectroscopy (fNIRS). bNIRS has the unique capacity of monitoring brain haemodynamics/oxygenation (measuring oxygenated and deoxygenated haemoglobin), and metabolism (measuring the changes in the redox state of cytochrome-c-oxidase). When combined with electroencephalography (EEG), bNIRS provides a unique neuromonitoring platform to explore neurovascular coupling mechanisms. In this paper, we present a novel pipeline for the integrated analysis of bNIRS and EEG signals, and demonstrate its use on multi-channel bNIRS data recorded with concurrent EEG on healthy adults during a visual stimulation task. We introduce the use of the Finite Impulse Response functions within the General Linear Model for bNIRS and show its feasibility to statistically localize the haemodynamic and metabolic activity in the occipital cortex. Moreover, our results suggest that the fusion of haemodynamic and metabolic measures unveils additional information on brain functioning over haemodynamic imaging alone. The cross-correlation-based analysis of interrelationships between electrical (EEG) and haemodynamic/metabolic (bNIRS) activity revealed that the bNIRS metabolic signal offers a unique marker of brain activity, being more closely coupled to the neuronal EEG response.

Sailing in rough waters: Examining volatility of fMRI noise

Background

The assumption that functional magnetic resonance imaging (fMRI) noise has constant volatility has recently been challenged by studies examining heteroscedasticity arising from head motion and physiological noise. The present study builds on this work using latest methods from the field of financial mathematics to model fMRI noise volatility.

Methods

Multi-echo phantom and human fMRI scans were used and realised volatility was estimated. The Hurst parameter H ∈ (0, 1), which governs the roughness/irregularity of realised volatility time series, was estimated. Calibration of H was performed pathwise, using well-established neural network calibration tools.

Results

In all experiments the volatility calibrated to values within the rough case, H < 0.5, and on average fMRI noise was very rough with 0.03 < H < 0.05. Some edge effects were also observed, whereby H was larger near the edges of the phantoms.

Discussion

The findings suggest that fMRI volatility is not only non-constant, but also substantially more irregular than a standard Brownian motion. Thus, further research is needed to examine the impact such pronounced oscillations in the volatility of fMRI noise have on data analyses.