Functional Connectivity Graphs Research Articles

Exploring Implicit Biological Heterogeneity in ASD Diagnosis Using a Multi-Head Attention Graph Neural Network.

Autism spectrum disorder (ASD) is a neurodevelopmental disorder exhibiting heterogeneous characteristics in patients, including variability in developmental progression and distinct neuroanatomical features influenced by sex and age. Recent advances in deep learning models based on functional connectivity (FC) graphs have produced promising results, but they have focused on generalized global activation patterns and failed to capture specialized regional characteristics and accurately assess disease indications. To overcome these limitations, we propose a novel deep learning method that models FC with multi-head attention, which enables simultaneous modeling of the intricate and variable patterns of brain connectivity associated with ASD, effectively extracting abnormal patterns of brain connectivity. The proposed method not only identifies region-specific correlations but also emphasizes connections at specific, transient time points from diverse perspectives. The extracted FC is transformed into a graph, assigning weighted labels to the edges to reflect the degree of correlation, which is then processed using a graph neural network capable of handling edge labels. Experiments on the autism brain imaging data exchange (ABIDE) I and II datasets, which include a heterogeneous cohort, showed superior performance over the state-of-the-art methods, improving accuracy by up to 3.7%p. The incorporation of multi-head attention in FC analysis markedly improved the distinction between typical brains and those affected by ASD. Additionally, the ablation study validated diverse brain characteristics in ASD patients across different ages and sexes, offering insightful interpretations. These results emphasize the effectiveness of the method in enhancing diagnostic accuracy and its potential in advancing neurological research for ASD diagnosis.

Exploring the Neural Processes behind Narrative Engagement: An EEG Study.

Past cognitive neuroscience studies using naturalistic stimuli have considered narratives holistically and focused on cognitive processes. In this study, we incorporated the narrative structure, the dramatic arc, as an object of investigation, to examine how engagement levels fluctuate across a narrative-aligned dramatic arc. We explored the possibility of predicting self-reported engagement ratings from neural activity and investigated the idiosyncratic effects of each phase of the dramatic arc on brain responses as well as the relationship between engagement and brain responses. We presented a movie excerpt following the six-phase narrative arc structure to female and male participants while collecting EEG signals. We then asked this group of participants to recall the excerpt, another group to segment the video based on the dramatic arc model, and a third to rate their engagement levels while watching the movie. The results showed that the self-reported engagement ratings followed the pattern of the narrative dramatic arc. Moreover, while EEG amplitude could not predict group-averaged engagement ratings, other features comprising dynamic intersubject correlation (dISC), including certain frequency bands, dynamic functional connectivity patterns and graph features were able to achieve this. Furthermore, neural activity in the last two phases of the dramatic arc significantly predicted engagement patterns. This study is the first to explore the cognitive processes behind the dramatic arc and its phases. By demonstrating how neural activity predicts self-reported engagement, which itself aligns with the narrative structure, this study provides insights on the interrelationships between narrative structure, neural responses, and viewer engagement.

Graph analysis of resting state functional brain networks and associations with cognitive outcomes in survivors of pediatric brain tumor

Survivors of pediatric brain tumors often live with long-term cognitive difficulties related to brain changes associated with the tumor itself as well as treatments such as radiation therapy. The present study used graph theory to examine functional network properties in this population and whether graph metrics relate to core cognitive skills: attention, working memory, and processing speed. 31 survivors and 31 matched controls completed neuropsychological testing and functional magnetic resonance imaging. Neuroimaging was preprocessed and spatially constrained ICA was completed, followed by calculation of area under the curve values of graph metrics. Results revealed a significant difference such that brain tumor survivors exhibited less small-world properties. This was found to be related to working memory, such that less small-worldness in the network was related to poorer performance. Furthermore, hub regions appear to be particularly vulnerable to disruption. Comparison to results of microstructural network analysis from a similar sample suggest functional connectivity graph metrics provide different and complementary information and additional post-hoc analyses are also discussed. These findings reveal that survivors of pediatric brain tumor indeed display significant differences in functional brain networks that are quantifiable by graph theory and build a foundation to better understand how metrics such as small-worldness can be used to predict long-term cognitive outcomes in adulthood. Ongoing neuroimaging research may play a part in precision medicine determining treatment protocols and interventions for pediatric brain tumor patients.

Reliable evaluation of functional connectivity and graph theory measures in source-level EEG: How many electrodes are enough?

ObjectiveUsing EEG to characterise functional brain networks through graph theory has gained significant interest in clinical and basic research. However, the minimal requirements for reliable measures remain largely unaddressed. Here, we examined functional connectivity estimates and graph theory metrics obtained from EEG with varying electrode densities. MethodsEEG was recorded with 128 electrodes in 33 participants. The high-density EEG data were subsequently subsampled into three sparser montages (64, 32, and 19 electrodes). Four inverse solutions, four measures of functional connectivity, and five graph theory metrics were tested. ResultsThe correlation between the results obtained with 128-electrode and the subsampled montages decreased as a function of the number of electrodes. As a result of decreased electrode density, the network metrics became skewed: mean network strength and clustering coefficient were overestimated, while characteristic path length was underestimated. ConclusionsSeveral graph theory metrics were altered when electrode density was reduced. Our results suggest that, for optimal balance between resource demand and result precision, a minimum of 64 electrodes should be utilised when graph theory metrics are used to characterise functional brain networks in source-reconstructed EEG data. SignificanceCharacterisation of functional brain networks derived from low-density EEG warrants careful consideration.

Automatic feature learning model combining functional connectivity network and graph regularization for depression detection

Depression has become a major health and economic burden worldwide. Electroencephalography (EEG) data has been used by a growing number of researchers to study depression. EEG-based functional connectivity (FC) features have emerged since they can account for the relationships between different brain regions. In this paper, the time–frequency analysis technique is introduced into the construction of the FC matrix. Specifically, instead of directly building the FC matrix from the EEG signals, the intrinsic time-scale decomposition (ITD) method is employed to mine the time–frequency information, and then the Pearson correlation is used to measure the FC between channels. The results show the significant differences in the FC networks between different groups. Furthermore, the graph-based adaptive least absolute shrinkage and selection operator model (GA-LASSO) is proposed in this paper to learn the discriminative features from the FC matrix, which is mainly achieved by adding both the adaptive L1 and graph regularized terms to the original least absolute shrinkage and selection operator (LASSO) model. The advantages of GA-LASSO come from the processing of discriminative weights of different features, and the connections between features by graph topology. In addition, the effectiveness of the proposed strategy of depression detection is validated on the open dataset MODMA, as well as the self-collected dataset called EDRA. The experimental results show that the current study sheds new light on the pathological mechanism of subclinical depression and suggests that EEG resting-state FC analysis may identify potentially effective biomarkers for its clinical diagnosis.

Effective Connectivity Is A More Promising Biomarker for Brain Health in Middle‐Aged Adults Than Functional Connectivity: Bogalusa Heart Study

AbstractBackgroundEffective connectivity, the causal influence that functional activity in a source brain location exerts over functional activity in a target brain location, has the potential to provide richer information about brain networks than functional connectivity, which only quantifies activity synchrony between regions. However, head‐to‐head comparisons of effective and functional connectivity from functional MRI data are rare, especially among diverse cognitively healthy middle aged and older adults.Method100 Bogalusa Heart Study participants performed a Stroop task during functional MRI (Table 1). Effective connectivity and functional connectivity among 24 regions of interest (ROIs) previously identified as activated by this task [1] were calculated using deep stacking networks [2] and Pearson correlation, respectively. Graph metrics were then calculated from directed graphs resulting from effective connectivity and undirected graphs resulting from functional connectivity. Linear regression models related graph metrics to demographic, cardiometabolic, and cognitive measures.ResultFigure 1 shows the group‐consensus effective and functional connectivity graphs. Systolic blood pressure (SBP) was associated with effective connectivity based characteristic path length (Pearson’s r = 0.22, p‐value = 0.038) and modularity (r = 0.25, p‐value = 0.02) (Figure 2). Diastolic blood pressure (DBP) was associated with effective connectivity based degree, density, clustering coefficient, modularity, transitivity, characteristic path length, and global efficiency (|r| = 0.26 ‐ 0.35, maximum p‐value = 0.013) (Figure 2). Word reading and vocabulary scores from the Wide Range Achievement Test (WRAT) were associated with effective connectivity based degree, density, clustering coefficient, transitivity, characteristic path length, modularity, and global efficiency (|r| = 0.21 ‐ 0.27, maximum p‐value = 0.049). Logical Memory II Recognition scores associated with effective connectivity based assortativity (r = 0.24, p‐value = 0.025). Associations between functional connectivity based graph metrics and demographic, cardiometabolic, and cognitive measures were not significant.ConclusionIn a diverse, cognitively healthy, middle‐aged community sample, graph metrics derived from effective connectivity based directed graphs tracked more closely with recognized indicators of brain health than traditional functional connectivity based undirected graph metrics.

Effective Connectivity Is A More Promising Biomarker for Brain Health in Middle‐Aged Adults Than Functional Connectivity: Bogalusa Heart Study

AbstractBackgroundEffective connectivity, the causal influence that functional activity in a source brain location exerts over functional activity in a target brain location, has the potential to provide richer information about brain networks than functional connectivity, which only quantifies activity synchrony between regions. However, head‐to‐head comparisons of effective and functional connectivity from functional MRI data are rare, especially among diverse cognitively healthy middle aged and older adults.Method100 Bogalusa Heart Study participants performed a Stroop task during functional MRI (Table 1). Effective connectivity and functional connectivity among 24 regions of interest (ROIs) previously identified as activated by this task [1] were calculated using deep stacking networks [2] and Pearson correlation, respectively. Graph metrics were then calculated from directed graphs resulting from effective connectivity and undirected graphs resulting from functional connectivity. Linear regression models related graph metrics to demographic, cardiometabolic, and cognitive measures.ResultFigure 1 shows the group‐consensus effective and functional connectivity graphs. Systolic blood pressure (SBP) was associated with effective connectivity based characteristic path length (Pearson’s r = 0.22, p‐value = 0.038) and modularity (r = 0.25, p‐value = 0.02) (Figure 2). Diastolic blood pressure (DBP) was associated with effective connectivity based degree, density, clustering coefficient, modularity, transitivity, characteristic path length, and global efficiency (|r| = 0.26 ‐ 0.35, maximum p‐value = 0.013) (Figure 2). Word reading and vocabulary scores from the Wide Range Achievement Test (WRAT) were associated with effective connectivity based degree, density, clustering coefficient, transitivity, characteristic path length, modularity, and global efficiency (|r| = 0.21 ‐ 0.27, maximum p‐value = 0.049). Logical Memory II Recognition scores associated with effective connectivity based assortativity (r = 0.24, p‐value = 0.025). Associations between functional connectivity based graph metrics and demographic, cardiometabolic, and cognitive measures were not significant.ConclusionIn a diverse, cognitively healthy, middle‐aged community sample, graph metrics derived from effective connectivity based directed graphs tracked more closely with recognized indicators of brain health than traditional functional connectivity based undirected graph metrics.

Diagnosis of Epilepsy with Functional Connectivity in EEG after a Suspected First Seizure.

Epilepsy is regarded as a structural and functional network disorder, affecting around 50 million people worldwide. A correct disease diagnosis can lead to quicker medical action, preventing adverse effects. This paper reports the design of a classifier for epilepsy diagnosis in patients after a first ictal episode, using electroencephalogram (EEG) recordings. The dataset consists of resting-state EEG from 629 patients, of which 504 were retained for the study. The patient's cohort exists out of 291 patients with epilepsy and 213 patients with other pathologies. The data were split into two sets: 80% training set and 20% test set. The extracted features from EEG included functional connectivity measures, graph measures, band powers and brain asymmetry ratios. Feature reduction was performed, and the models were trained using Machine Learning (ML) techniques. The models' evaluation was performed with the area under the receiver operating characteristic curve (AUC). When focusing specifically on focal lesional epileptic patients, better results were obtained. This classification task was optimized using a 5-fold cross-validation, where SVM using PCA for feature reduction achieved an AUC of 0.730 ± 0.030. In the test set, the same model achieved 0.649 of AUC. The verified decrease is justified by the considerable diversity of pathologies in the cohort. An analysis of the selected features across tested models shows that functional connectivity and its graph measures have the most considerable predictive power, along with full-spectrum frequency-based features. To conclude, the proposed algorithms, with some refinement, can be of added value for doctors diagnosing epilepsy from EEG recordings after a suspected first seizure.

Graph-generative neural network for EEG-based epileptic seizure detection via discovery of dynamic brain functional connectivity

Dynamic complexity in brain functional connectivity has hindered the effective use of signal processing or machine learning methods to diagnose neurological disorders such as epilepsy. This paper proposed a new graph-generative neural network (GGN) model for the dynamic discovery of brain functional connectivity via deep analysis of scalp electroencephalogram (EEG) signals recorded from various regions of a patient’s scalp. Brain functional connectivity graphs are generated for the extraction of spatial–temporal resolution of various onset epilepsy seizure patterns. Our supervised GGN model was substantiated by seizure detection and classification experiments. We train the GGN model using a clinically proven dataset of over 3047 epileptic seizure cases. The GGN model achieved a 91% accuracy in classifying seven types of epileptic seizure attacks, which outperformed the 65%, 74%, and 82% accuracy in using the convolutional neural network (CNN), graph neural networks (GNN), and transformer models, respectively. We present the GGN model architecture and operational steps to assist neuroscientists or brain specialists in using dynamic functional connectivity information to detect neurological disorders. Furthermore, we suggest to merge our spatial–temporal graph generator design in upgrading the conventional CNN and GNN models with dynamic convolutional kernels for accuracy enhancement.

The thresholding problem and variability in the EEG graph network parameters

Graph thresholding is a frequently used practice of eliminating the weak connections in brain functional connectivity graphs. The main aim of the procedure is to delete the spurious connections in the data. However, the choice of the threshold is arbitrary, and the effect of the threshold choice is not fully understood. Here we present the description of the changes in the global measures of a functional connectivity graph depending on the different proportional thresholds based on the 146 resting-state EEG recordings. The dynamics is presented in five different synchronization measures (wPLI, ImCoh, Coherence, ciPLV, PPC) in sensors and source spaces. The analysis shows significant changes in the graph’s global connectivity measures as a function of the chosen threshold which may influence the outcome of the study. The choice of the threshold could lead to different study conclusions; thus it is necessary to improve the reasoning behind the choice of the different analytic options and consider the adoption of different analytic approaches. We also proposed some ways of improving the procedure of thresholding in functional connectivity research.

A Novel Classification Framework Using the Graph Representations of Electroencephalogram for Motor Imagery Based Brain-Computer Interface

The motor imagery (MI) based brain-computer interfaces (BCIs) have been proposed as a potential physical rehabilitation technology. However, the low classification accuracy achievable with MI tasks is still a challenge when building effective BCI systems. We propose a novel MI classification model based on measurement of functional connectivity between brain regions and graph theory. Specifically, motifs describing local network structures in the brain are extracted from functional connectivity graphs. A graph embedding model called Ego-CNNs is then used to build a classifier, which can convert the graph from a structural representation to a fixed-dimensional vector for detecting critical structure in the graph. We validate our proposed method on four datasets, and the results show that our proposed method produces high classification accuracies in two-class classification tasks (92.8% for dataset 1, 93.4% for dataset 2, 96.5% for dataset 3, and 80.2% for dataset 4) and multiclass classification tasks (90.33% for dataset 1). Our proposed method achieves a mean Kappa value of 0.88 across nine participants, which is superior to other methods we compared it to. These results indicate that there is a local structural difference in functional connectivity graphs extracted under different motor imagery tasks. Our proposed method has great potential for motor imagery classification in future studies.

Greek High Phenolic Early Harvest Extra Virgin Olive Oil Reduces the Over-Excitation of Information-Flow Based on Dominant Coupling Mode (DoCM) Model in Patients with Mild Cognitive Impairment: An EEG Resting-State Validation Approach.

Extra virgin olive oil (EVOO) constitutes a natural compound with high protection over cognitive function that could positively alter brain dynamics and the mixture of within and between-frequency connectivity. The balance of cross-frequency coupling over within-frequency coupling can build a nonlinearity index (NI) that encapsulates the over-excitation of information flow between brain areas and across experimental time. The present study investigated for the very first time how the Greek High Phenolic Early Harvest Extra Virgin Olive Oil (HP-EH-EVOO) versus Moderate Phenolic (MP-EVOO) and Mediterranean Diet (MeDi) intervention in people with mild cognitive impairment (MCI) could affect their spontaneous EEG dynamic connectivity. Forty-three subjects (14 in MeDi, 16 in MP-EVOO, and 13 in HP-EH-EVOO) followed an EEG resting-state recording session (eyes-open and closed) before and after the treatment. Following our dominant coupling mode model, we built a dynamic integrated dynamic functional connectivity graph that tabulates the functional strength and the dominant coupling mode model of every pair of brain areas. Signal spectrum within 1-13 Hz and theta/beta ratio have decreased in the HP-EH-EVOO group in the eyes-open condition. The intervention improved the FIDoCM across groups and conditions but was more prominent in the HP-EH-EVOO group (p < 0.001). Finally, we revealed a significant higher post-intervention reduction of NI (ΔNITotal and α) for the HP-EH-EVOO compared to the MP-EVOO and MeDi groups (p < 0.0001). Long-term intervention with HP-EH-EVOO reduced the over-excitation of information flow in spontaneous brain activity and altered the signal spectrum of EEG rhythms.

Quantification of pain severity using EEG-based functional connectivity

Background and objectivesThe traditional pain measures are qualitative and inaccurate. Therefore, electroencephalography (EEG) signals have been recently used and analyzed to differentiate pain from no-pain state. The challenge is emerged when the accuracy of these classifiers is not enough for differentiating between different pain levels. In this paper, we demonstrate that EEG-based functional connectivity graph is remarkably changed by increasing the pain intensity and therefore, by deriving informative features from this graph at each pain level, we finely differentiate between five levels of pain. MethodsIn this research, 23 subjects (mean age: 22 years, Std: 1.4) are voluntarily enrolled and their EEG signals are recorded via 29 electrodes, while they execute the cold pressure test. The signals are recorded two times from each case, while subjects press a button to annotate the EEGs into five pain levels. After denoising the EEGs, the brain connectivity graph in the Alpha band are estimated using partial directed coherence method in successive time frames. By observing the differences of connectivity graph features in different levels of pain, a bio-inspired decision tree (multilayer support vector machines) is proposed. Discriminant features are selected using sequential forward feature selection manner and the selected features are applied to the proposed decision tree. ResultsClassification result for differentiating between the pain and no-pain states provides 92% accuracy (94% sensitivity and 91% specificity), while for the five classes of pain, the proposed scheme generates 86% accuracy (90% sensitivity and 82% specificity), which is slightly decreased compared to the two-class condition. Moreover, the results are evaluated in terms of robustness against noise in different signal to noise ratio levels. Comparison results with previous research imply the significant superiority of the proposed scheme. ConclusionIn this paper, we show that the elicited features from the filtered brain graph are able to significantly discriminate five different levels of pain. This is therefore the amount of co-activation between the brain regions (graph links) are significantly varied, as the pain feeling increases. Our observations are consistent with the physiological observations acquired from the images of functional magnetic resonance and magnetoencephalography.

Early classification of motor tasks using dynamic functional connectivity graphs from EEG

Objective. Classification of electroencephalography (EEG) signals with high accuracy using short recording intervals has been a challenging problem in developing brain computer interfaces (BCIs). This paper presents a novel feature extraction method for EEG recordings to tackle this problem. Approach. The proposed approach is based on the concept that the brain functions in a dynamic manner, and utilizes dynamic functional connectivity graphs. The EEG data is first segmented into intervals during which functional networks sustain their connectivity. Functional connectivity networks for each identified segment are then localized, and graphs are constructed, which will be used as features. To take advantage of the dynamic nature of the generated graphs, a long short term memory classifier is employed for classification. Main results. Features extracted from various durations of post-stimulus EEG data associated with motor execution and imagery tasks are used to test the performance of the classifier. Results show an average accuracy of 85.32% about only 500 ms after stimulus presentation. Significance. Our results demonstrate, for the first time, that using the proposed feature extraction method, it is possible to classify motor tasks from EEG recordings using a short interval of the data in the order of hundreds of milliseconds (e.g. 500 ms). This duration is considerably shorter than what has been reported before. These results will have significant implications for improving the effectiveness and the speed of BCIs, particularly for those used in assistive technologies.

Identifying Mild Hepatic Encephalopathy Based on Multi-Layer Modular Algorithm and Machine Learning.

Hepatic encephalopathy (HE) is a neurocognitive dysfunction based on metabolic disorders caused by severe liver disease, which has a high one-year mortality. Mild hepatic encephalopathy (MHE) has a high risk of converting to overt HE, and thus the accurate identification of MHE from cirrhosis with no HE (noHE) is of great significance in reducing mortality. Previously, most studies focused on studying abnormality in the static brain networks of MHE to find biomarkers. In this study, we aimed to use multi-layer modular algorithm to study abnormality in dynamic graph properties of brain network in MHE patients and construct a machine learning model to identify individual MHE from noHE. Here, a time length of 500-second resting-state functional MRI data were collected from 41 healthy subjects, 32 noHE patients and 30 MHE patients. Multi-layer modular algorithm was performed on dynamic brain functional connectivity graph. The connection-stability score was used to characterize the loyalty in each brain network module. Nodal flexibility, cohesion and disjointness were calculated to describe how the node changes the network affiliation across time. Results show that significant differences between MHE and noHE were found merely in nodal disjointness in higher cognitive network modules (ventral attention, fronto-parietal, default mode networks) and these abnormalities were associated with the decline in patients’ attention and visual memory function evaluated by Digit Symbol Test. Finally, feature extraction from node disjointness with the support vector machine classifier showed an accuracy of 88.71% in discrimination of MHE from noHE, which was verified by different window sizes, modular partition parameters and machine learning parameters. All these results show that abnormal nodal disjointness in higher cognitive networks during brain network evolution can be seemed as a biomarker for identification of MHE, which help us understand the disease mechanism of MHE at a fine scale.

Neural functional connectivity in patients with periventricular nodular heterotopia-mediated epilepsy

Periventricular nodular heterotopia (PNH) is characterized by disabled neural migration and is usually associated with epilepsy. Despite awareness of PNH-related epilepsy, little is known about the brain-level underlying functional neural bases. Thus, we used functional magnetic resonance imaging (MRI) to examine the neurobiology of 42 subjects with PNH-related epilepsy and 42 sex- and age-matched healthy controls. Measurements of functional connectivity (FC) and whole-brain graph theory analysis of data in the resting state were performed to assess neurological organization and topology. PNH patients exhibited significantly higher FC in the parietal lobe, cingulum and thalamus, as well as significantly lower FC in frontoparietal, hippocampal, and precentral regions. Graph theory analysis identified no significant differences between patients and controls, while patients showed lower network global efficiency in the limbic and cerebellum network and occipital cortex. Seed-based FC analysis confirmed disruption of activities and interregional connectivity in remote epileptic networks of patients, which may point to underlying pathological mechanisms. The cerebellum and limbic system of patients showed altered topology, suggesting that these regions or hubs may contribute to whole-brain circuits in PNH and epilepsy.

Intrinsic Network Activity in Human Brain Organoids

Human brain organoids replicate much of the cellular diversity and developmental anatomy of the human brain. However, neurophysiology within organoids remains relatively under-explored. With high-density CMOS microelectrode arrays and shank electrodes, we probed broadband and three-dimensional activity of human brain organoids. From spiking activity, we estimated a directed functional connectivity graph of neural network activity which showed a large number of weak functional connections enmeshed within a network skeleton of significantly fewer strong connections. Increasing the intrinsic inhibitory tone with a benzodiazepine suppressed the network skeleton of the functional network. We further demonstrate that human brain organoids have self-organized neuronal assemblies of sufficient size, cellular orientation, and functional connectivity to generate local field potentials from their collective transmembrane currents that phase-lock to spiking activity. These results point to the potential of brain organoids for the study of neuropsychiatric diseases, drug mechanisms, and the effects of external stimuli upon neuronal networks.

Quantitative Identification of Functional Connectivity Disturbances in Neuropsychiatric Lupus Based on Resting-State fMRI: A Robust Machine Learning Approach.

Neuropsychiatric systemic lupus erythematosus (NPSLE) is an autoimmune entity comprised of heterogenous syndromes affecting both the peripheral and central nervous system. Research on the pathophysiological substrate of NPSLE manifestations, including functional neuroimaging studies, is extremely limited. The present study examined person-specific patterns of whole-brain functional connectivity in NPSLE patients (n = 44) and age-matched healthy control participants (n = 39). Static functional connectivity graphs were calculated comprised of connection strengths between 90 brain regions. These connections were subsequently filtered through rigorous surrogate analysis, a technique borrowed from physics, novel to neuroimaging. Next, global as well as nodal network metrics were estimated for each individual functional brain network and were input to a robust machine learning algorithm consisting of a random forest feature selection and nested cross-validation strategy. The proposed pipeline is data-driven in its entirety, and several tests were performed in order to ensure model robustness. The best-fitting model utilizing nodal graph metrics for 11 brain regions was associated with 73.5% accuracy (74.5% sensitivity and 73% specificity) in discriminating NPSLE from healthy individuals with adequate statistical power. Closer inspection of graph metric values suggested an increased role within the functional brain network in NSPLE (indicated by higher nodal degree, local efficiency, betweenness centrality, or eigenvalue efficiency) as compared to healthy controls for seven brain regions and a reduced role for four areas. These findings corroborate earlier work regarding hemodynamic disturbances in these brain regions in NPSLE. The validity of the results is further supported by significant associations of certain selected graph metrics with accumulated organ damage incurred by lupus, with visuomotor performance and mental flexibility scores obtained independently from NPSLE patients.

Nonparametric Bayesian estimation for multivariate Hawkes processes

This paper studies nonparametric estimation of parameters of multivariate Hawkes processes. We consider the Bayesian setting and derive posterior concentration rates. First, rates are derived for $\mathbb{L}_{1}$-metrics for stochastic intensities of the Hawkes process. We then deduce rates for the $\mathbb{L}_{1}$-norm of interactions functions of the process. Our results are exemplified by using priors based on piecewise constant functions, with regular or random partitions and priors based on mixtures of Betas distributions. We also present a simulation study to illustrate our results and to study empirically the inference on functional connectivity graphs of neurons

Multiclass classification of patients during different stages of Alzheimer’s disease using fMRI time-series

Alzheimer’s Disease (AD) begins several years before the symptoms develop. It starts with Mild Cognitive Impairment (MCI) which can be separated into Early MCI and Late MCI (EMCI and LMCI). Functional connectivity analysis and classification are done among the different stages of illness with Functional Magnetic Resonance Imaging (fMRI). In this study, in addition to the four stages including healthy, EMCI, LMCI, and AD, the patients have been tracked for a year. Indeed, the classification has been done among 7 groups to analyze the functional connectivity changes in one year in different stages. After generating the functional connectivity graphs for eliminating the weak links, three different sparsification methods were used. In addition to simple thresholding, spectral sparsification based on effective resistance and sparse autoencoder were performed in order to analyze the effect of sparsification routine on classification results. Also, instead of extracting common features, the correlation matrices were reshaped to a correlation vector and used as a feature vector to enter the classifier. Since the correlation matrix is symmetric, in another analysis half of the feature vector was used, moreover, the Genetic Algorithm (GA) also utilized for feature vector dimension reduction. The non-linear SVM classifier with a polynomial kernel applied. The results showed that the autoencoder sparsification method had the greatest discrimination power with the accuracy of 98.35% for classification when the feature vector was the full correlation matrix.