Sparsity-guided multiple functional connectivity patterns for classification of schizophrenia via convolutional network.

The explorations of brain functional connectivity network (FCN) using resting-state functional magnetic resonance imaging can provide crucial insights into discriminative analysis of neuropsychiatric disorders, such as schizophrenia (SZ). Pearson's correlation (PC) is widely used to construct a densely connected FCN which may overlook some complex interactions of paired regions of interest (ROIs) under confounding effect of other ROIs. Although the method of sparse representation takes into account this issue, it penalizes each edge equally, which often makes the FCN look like a random network. In this paper, we establish a new framework, called convolutional neural network with sparsity-guided multiple functional connectivity, for SZ classification. The framework consists of two components. (1) The first component constructs a sparse FCN by integrating PC and weighted sparse representation (WSR). The FCN retains the intrinsic correlation between paired ROIs, and eliminates false connection simultaneously, resulting in sparse interactions among multiple ROIs with the confounding effect regressed out. (2) In the second component, we develop a functional connectivity convolution to learn discriminative features for SZ classification from multiple FCNs by mining the joint spatial mapping of FCNs. Finally, an occlusion strategy is employed to explore the contributive regions and connections, to derive the potential biomarkers in identifying associated aberrant connectivity of SZ. The experiments on SZ identification verify the rationality and advantages of our proposed method. This framework also can be used as a diagnostic tool for other neuropsychiatric disorders.

[Alterations of ß-γ coupling of scalp electroencephalography during epilepsy].

Uncovering the alterations of neural interactions within the brain during epilepsy is important for the clinical diagnosis and treatment. Previous studies have shown that the phase-amplitude coupling (PAC) can be used as a potential biomarker for locating epileptic zones and characterizing the transition of epileptic phases. However, in contrast to the θ-γ coupling widely investigated in epilepsy, few studies have paid attention to the ß-γ coupling, as well as its potential applications. In the current study, we use the modulation index (MI) to calculate the scalp electroencephalography (EEG)-based ß-γ coupling and investigate the corresponding changes during different epileptic phases. The results show that the ß-γ coupling of each brain region changes with the evolution of epilepsy, and in several brain regions, the ß-γ coupling decreases during the ictal period but increases in the post-ictal period, where the differences are statistically significant. Moreover, the alterations of ß-γ coupling between different brain regions can also be observed, and the strength of ß-γ coupling increases in the post-ictal period, where the differences are also significant. Taken together, these findings not only contribute to understanding neural interactions within the brain during the evolution of epilepsy, but also provide a new insight into the clinical treatment.

A toolbox for brain network construction and classification (BrainNetClass).

Brain functional network has been increasingly used in understanding brain functions and diseases. While many network construction methods have been proposed, the progress in the field still largely relies on static pairwise Pearson's correlation-based functional network and group-level comparisons. We introduce a "Brain Network Construction and Classification (BrainNetClass)" toolbox to promote more advanced brain network construction methods to the filed, including some state-of-the-art methods that were recently developed to capture complex and high-order interactions among brain regions. The toolbox also integrates a well-accepted and rigorous classification framework based on brain connectome features toward individualized disease diagnosis in a hope that the advanced network modeling could boost the subsequent classification. BrainNetClass is a MATLAB-based, open-source, cross-platform toolbox with both graphical user-friendly interfaces and a command line mode targeting cognitive neuroscientists and clinicians for promoting reliability, reproducibility, and interpretability of connectome-based, computer-aided diagnosis. It generates abundant classification-related results from network presentations to contributing features that have been largely ignored by most studies to grant users the ability of evaluating the disease diagnostic model and its robustness and generalizability. We demonstrate the effectiveness of the toolbox on real resting-state functional MRI datasets. BrainNetClass (v1.0) is available at https://github.com/zzstefan/BrainNetClass.

[Research on brain network for schizophrenia classification based on resting-state functional magnetic resonance imaging].

How to extract high discriminative features that help classification from complex resting-state fMRI (rs-fMRI) data is the key to improving the accuracy of brain disease recognition such as schizophrenia. In this work, we use a weighted sparse model for brain network construction, and utilize the Kendall correlation coefficient (KCC) to extract the discriminative connectivity features for schizophrenia classification, which is conducted with the linear support vector machine. Experimental results based on the rs-fMRI of 57 schizophrenia patients and 64 healthy controls show that our proposed method is more effective ( i.e., achieving a significantly higher classification accuracy, 81.82%) than other competing methods. Specifically, compared with the traditional network construction methods (Pearson's correlation and sparse representation) and the commonly used feature selection methods (two-sample t-test and Least absolute shrinkage and selection operator (Lasso)), the algorithm proposed in this paper can more effectively extract the discriminative connectivity features between the schizophrenia patients and the healthy controls, and further improve the classification accuracy. At the same time, the discriminative connectivity features extracted in the work could be used as the potential clinical biomarkers to assist the identification of schizophrenia.

Weighted Graph Regularized Sparse Brain Network Construction for MCI Identification.

Brain functional networks (BFNs) constructed from resting-state functional magnetic resonance imaging (rs-fMRI) have been widely applied to the analysis and diagnosis of brain diseases, such as Alzheimer's disease and its prodrome, namely mild cognitive impairment (MCI). Constructing a meaningful brain network based on, for example, sparse representation (SR) is the most essential step prior to the subsequent analysis or disease identification. However, the independent coding process of SR fails to capture the intrinsic locality and similarity characteristics in the data. To address this problem, we propose a novel weighted graph (Laplacian) regularized SR framework, based on which BFN can be optimized by considering both intrinsic correlation similarity and local manifold structure in the data, as well as sparsity prior of the brain connectivity. Additionally, the non-convergence of the graph Laplacian in the self-representation model has been solved properly. Combined with a pipeline of sparse feature selection and classification, the effectiveness of our proposed method is demonstrated by identifying MCI based on the constructed BFNs.

Connectivity strength-weighted sparse group representation-based brain network construction for MCI classification.

Brain functional network analysis has shown great potential in understanding brain functions and also in identifying biomarkers for brain diseases, such as Alzheimer's disease (AD) and its early stage, mild cognitive impairment (MCI). In these applications, accurate construction of biologically meaningful brain network is critical. Sparse learning has been widely used for brain network construction; however, its l1 -norm penalty simply penalizes each edge of a brain network equally, without considering the original connectivity strength which is one of the most important inherent linkwise characters. Besides, based on the similarity of the linkwise connectivity, brain network shows prominent group structure (i.e., a set of edges sharing similar attributes). In this article, we propose a novel brain functional network modeling framework with a "connectivity strength-weighted sparse group constraint." In particular, the network modeling can be optimized by considering both raw connectivity strength and its group structure, without losing the merit of sparsity. Our proposed method is applied to MCI classification, a challenging task for early AD diagnosis. Experimental results based on the resting-state functional MRI, from 50 MCI patients and 49 healthy controls, show that our proposed method is more effective (i.e., achieving a significantly higher classification accuracy, 84.8%) than other competing methods (e.g., sparse representation, accuracy = 65.6%). Post hoc inspection of the informative features further shows more biologically meaningful brain functional connectivities obtained by our proposed method. Hum Brain Mapp 38:2370-2383, 2017. © 2017 Wiley Periodicals, Inc.

An EEG Study on [Formula: see text] Phase-Amplitude Coupling-Based Functional Brain Network in Epilepsy Patients.

Epilepsy, a chronic neuropsychiatric brain disorder characterized with recurrent seizures, is closely associated with abnormal neural communications within the brain. Despite that the phase-amplitude coupling (PAC) has been suggested to offer a new way to observe neural interactions during epilepsy, however, few studies pay attention to alterations of the epileptic functional brain network based on PAC, especially on the [Formula: see text] PAC. Therefore, we use scalp electroencephalography (EEG) data of epileptic patients and the [Formula: see text] PAC modulation index (MI) to construct functional brain networks to examine variations of neural interactions during different epileptic phases. Statistically, the findings show that between-channel MI values in the post-ictal period significantly increase compared to that in the pre-ictal period, and the between-channel MI value has a close association with the information of phase and amplitude provided by the channels. Importantly, in both the phase-amplitude and amplitude-phase functional brain networks, the average node degree is remarkably higher in the post-ictal period than that in the pre-ictal period, whereas the characteristic path length in the ictal and post-ictal periods is significantly lower than that in the pre-ictal period. Besides, the average betweenness centrality in the post-ictal period is remarkably higher than that in the ictal period. Interestingly, the positive correlations between within-channel MI values and between-channel MI values can be observed during the pre-ictal, ictal and post-ictal periods. These findings suggest that the [Formula: see text] PAC-based functional brain network may provide a novel perspective to understanding alterations of neural interactions during the epileptic evolution, and may contribute to effectively controlling the spread of epileptic seizures.

Dynamic graph transformer network via dual-view connectivity for autism spectrum disorder identification.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that requires objective and accurate identification methods for effective early intervention. Previous population-based methods via functional connectivity (FC) analysis ignore the differences between positive and negative FCs, which provide the potential information complementarity. And they also require additional information to construct a pre-defined graph. Meanwhile, two challenging demand attentions are the imbalance of performance caused by the class distribution and the inherent heterogeneity of multi-site data. In this paper, we propose a novel dynamic graph Transformer network based on dual-view connectivity for ASD Identification. It is based on the Autoencoders, which regard the input feature as individual feature and without any inductive bias. First, a dual-view feature extractor is designed to extract individual and complementary information from positive and negative connectivity. Then Graph Transformer network is innovated with a hot plugging K-Nearest Neighbor (KNN) algorithm module which constructs a dynamic population graph without any additional information. Additionally, we introduce the PolyLoss function and the Vrex method to address the class imbalance and improve the model's generalizability. The evaluation experiment on 1102 subjects from the ABIDE I dataset demonstrates our method can achieve superior performance over several state-of-the-art methods and satisfying generalizability for ASD identification.

Multi-Graph Attention Networks with Bilinear Convolution for Diagnosis of Schizophrenia.

The explorations of brain functional connectivity (FC) network using resting-state functional magnetic resonance imaging (rs-fMRI) can provide crucial insights into discriminative analysis of neuropsychiatric disorders such as schizophrenia (SZ). Graph attention network (GAT), which could capture the local stationary on the network topology and aggregate the features of neighboring nodes, has advantages in learning the feature representation of brain regions. However, GAT only can obtain the node-level features that reflect local information, ignoring the spatial information within the connectivity-based features that proved to be important for SZ diagnosis. In addition, existing graph learning techniques usually rely on a single graph topology to represent neighborhood information, and only consider a single correlation measure for connectivity features. Comprehensive analysis of multiple graph topologies and multiple measures of FC can leverage their complementary information that may contribute to identifying patients. In this paper, we propose a multi-graph attention network (MGAT) with bilinear convolution (BC) neural network framework for SZ diagnosis and functional connectivity analysis. Besides multiple correlation measures to construct connectivity networks from different perspectives, we further propose two different graph construction methods to capture both the low- and high-level graph topologies, respectively. Especially, the MGAT module is developed to learn multiple node interaction features on each graph topology, and the BC module is utilized to learn the spatial connectivity features of the brain network for disease prediction. Importantly, the rationality and advantages of our proposed method can be validated by the experiments on SZ identification. Therefore, we speculate that this framework may also be potentially used as a diagnostic tool for other neuropsychiatric disorders.

A model description of beta oscillations in the external globus pallidus.

The external globus pallidus (GPe), a subcortical nucleus located in the indirect pathway of the basal ganglia, is widely considered to have tight associations with abnormal beta oscillations (13-30 Hz) observed in Parkinson's disease (PD). Despite that many mechanisms have been put forward to explain the emergence of these beta oscillations, however, it is still unclear the functional contributions of the GPe, especially, whether the GPe itself can generate beta oscillations. To investigate the role played by the GPe in producing beta oscillations, we employ a well described firing rate model of the GPe neural population. Through extensive simulations, we find that the transmission delay within the GPe-GPe pathway contributes significantly to inducing beta oscillations, and the impacts of the time constant and connection strength of the GPe-GPe pathway on generating beta oscillations are non-negligible. Moreover, the GPe firing patterns can be significantly modulated by the time constant and connection strength of the GPe-GPe pathway, as well as the transmission delay within the GPe-GPe pathway. Interestingly, both increasing and decreasing the transmission delay can push the GPe firing pattern from beta oscillations to other firing patterns, including oscillation and non-oscillation firing patterns. These findings suggest that if the transmission delays within the GPe are at least 9.8 ms, beta oscillations can be produced originally in the GPe neural population, which also may be the origin of PD-related beta oscillations and should be regarded as a promising target for treatments for PD.

Outcome Prediction of Unconscious Patients Based on Weighted Sparse Brain Network Construction.

It is quite challenging to establish a prompt and reliable prognosis assessment for acquired brain injury (ABI) patients with persistent severe disorders of consciousness (DOC) like unconscious comatose and unresponsive wakefulness syndrome (a.k.a., vegetative state). Recent advances in brain functional imaging and functional net-work analysis have demonstrated its potential in determining the consciousness level and prognostic outcome for ABI patients with DOC. However, the diagnostic and prognostic usefulness of the whole-brain functional connectome based on advanced machine learning techniques has not been fully evaluated. The first aim of this study is to predict the outcome of individual unconscious ABI patients during a three-month follow-up. The second aim is to conduct precise individualized differentiation among different consciousness levels for exploring the neurobiological mechanisms underlying DOC. Based on resting-state fMRI, we construct large-scale functional networks by using a weighted sparse model, which ensures sparsity and interpretability by preserving strong functional connections. The functional connection strengths are exploited as features for outcome prediction and consciousness level differentiation. We achieve significantly improved consciousness level classification (accuracy: 84.78%) and recovery outcome prediction (accuracy: 89.74%) compared to other network construction methods. More importantly, we reveal the contributive connections across the entire brain in both tasks. These connections could serve as the potential biomarkers for better understanding of consciousness and further provide new insight into the development of diagnostic, prognostic, and effective therapeutic guidelines for ABI patients with DOC.

Temporal-spatial dynamic functional connectivity analysis in schizophrenia classification.

With the development of resting-state functional magnetic resonance imaging (rs-fMRI) technology, the functional connectivity network (FCN) which reflects the statistical similarity of temporal activity between brain regions has shown promising results for the identification of neuropsychiatric disorders. Alteration in FCN is believed to have the potential to locate biomarkers for classifying or predicting schizophrenia (SZ) from healthy control. However, the traditional FCN analysis with stationary assumption, i.e., static functional connectivity network (SFCN) at the time only measures the simple functional connectivity among brain regions, ignoring the dynamic changes of functional connectivity and the high-order dynamic interactions. In this article, the dynamic functional connectivity network (DFCN) is constructed to delineate the characteristic of connectivity variation across time. A high-order functional connectivity network (HFCN) designed based on DFCN, could characterize more complex spatial interactions across multiple brain regions with the potential to reflect complex functional segregation and integration. Specifically, the temporal variability and the high-order network topology features, which characterize the brain FCNs from region and connectivity aspects, are extracted from DFCN and HFCN, respectively. Experiment results on SZ identification prove that our method is more effective (i.e., obtaining a significantly higher classification accuracy, 81.82%) than other competing methods. Post hoc inspection of the informative features in the individualized classification task further could serve as the potential biomarkers for identifying associated aberrant connectivity in SZ.

Insights on the role of external globus pallidus in controlling absence seizures.

Absence epilepsy, characterized by transient loss of awareness and bilaterally synchronous 2-4 Hz spike and wave discharges (SWDs) on electroencephalography (EEG) during absence seizures, is generally believed to arise from abnormal interactions between the cerebral cortex (Ctx) and thalamus. Recent animal electrophysiological studies suggested that changing the neural activation level of the external globus pallidus (GPe) neurons can remarkably modify firing rates of the thalamic reticular nucleus (TRN) neurons through the GABAergic GPe-TRN pathway. However, the existing experimental evidence does not provide a clear answer as to whether the GPe-TRN pathway contributes to regulating absence seizures. Here, using a biophysically based mean-field model of the GPe-corticothalamic (GCT) network, we found that both directly decreasing the strength of the GPe-TRN pathway and inactivating GPe neurons can effectively suppress absence seizures. Also, the pallido-cortical pathway and the recurrent connection of GPe neurons facilitate the regulation of absence seizures through the GPe-TRN pathway. Specifically, in the controllable situation, enhancing the coupling strength of either of the two pathways can successfully terminate absence seizures. Moreover, the competition between the GPe-TRN and pallido-cortical pathways may lead to the GPe bidirectionally controlling absence seizures, and this bidirectional control manner can be significantly modulated by the Ctx-TRN pathway. Importantly, when the strength of the Ctx-TRN pathway is relatively strong, the bidirectional control of absence seizures by changing GPe neural activities can be observed at both weak and strong strengths of the pallido-cortical pathway.These findings suggest that the GPe-TRN pathway may have crucial functional roles in regulating absence seizures, which may provide a testable hypothesis for further experimental studies and new perspectives on the treatment of absence epilepsy.

Coarse Classification to Region-Scalable Refining for White Matter Lesions Segmentation in Multi-Channel MRI.

Brain lesions, especially White Matter Lesions (WMLs) that mostly found on magnetic resonance images of elderly people, are not only associated with normal aging, but also with various geriatric disorders including cardiovascular diseases, vascular disease, psychiatric disorders and dementia. Quantitative analysis of WMLs in large clinical trials is crucial in scientific investigations of such neurological diseases as well as in studying aging processes. Exploiting the different appearances of WMLs in multiple modalities, we propose a novel coarse classification to region-scalable refining method to segment WMLs in Magnetic Resonance Imaging (MRI) sequences without user intervention. Specifically, a nonlinear voxel-wise classifier is trained based on intensity features extracted from multimodality MRI sequences, and tissues' probabilistic prior provided by partial volume estimate images in native space. By considering the prior that the WMLs almost exist in white matter, a rejection algorithm is then used to eliminate the false-positive labels from the initial coarse classification. To further segment precise lesions boundary and detect missing lesions, a region-scalable refining is finally employed to effectively segment the WMLs based on the previous initial contour. Compared with the manual segmentation results from an experienced neuroradiologist, the segmentations for real images of our proposal show desirable performances and high accuracy and provide competitive solution with stateof- the-art methods.

Correlation-Weighted Sparse Group Representation for Brain Network Construction in MCI Classification.

Analysis of brain functional connectivity network (BFCN) has shown great potential in understanding brain functions and identifying biomarkers for neurological and psychiatric disorders, such as Alzheimer's disease and its early stage, mild cognitive impairment (MCI). In all these applications, the accurate construction of biologically meaningful brain network is critical. Due to the sparse nature of the brain network, sparse learning has been widely used for complex BFCN construction. However, the conventional l1-norm penalty in the sparse learning equally penalizes each edge (or link) of the brain network, which ignores the link strength and could remove strong links in the brain network. Besides, the conventional sparse regularization often overlooks group structure in the brain network, i.e., a set of links (or connections) sharing similar attribute. To address these issues, we propose to construct BFCN by integrating both link strength and group structure information. Specifically, a novel correlation-weighted sparse group constraint is devised to account for and balance among (1) sparsity, (2) link strength, and (3) group structure, in a unified framework. The proposed method is applied to MCI classification using the resting-state fMRI from ADNI-2 dataset. Experimental results show that our method is effective in modeling human brain connectomics, as demonstrated by superior MCI classification accuracy of 81.8%. Moreover, our method is promising for its capability in modeling more biologically meaningful sparse brain networks, which will benefit both basic and clinical neuroscience studies.

Learning-based 3T brain MRI segmentation with guidance from 7T MRI labeling.

PURPOSE: Segmentation of brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is crucial for brain structural measurement and disease diagnosis. Learning-based segmentation methods depend largely on the availability of good training ground truth. However, the commonly used 3T MR images are of insufficient image quality and often exhibit poor intensity contrast between WM, GM, and CSF. Therefore, they are not ideal for providing good ground truth label data for training learning-based methods. Recent advances in ultrahigh field 7T imaging make it possible to acquire images with excellent intensity contrast and signal-to-noise ratio. METHODS: In this paper, the authors propose an algorithm based on random forest for segmenting 3T MR images by training a series of classifiers based on reliable labels obtained semiautomatically from 7T MR images. The proposed algorithm iteratively refines the probability maps of WM, GM, and CSF via a cascade of random forest classifiers for improved tissue segmentation. RESULTS: The proposed method was validated on two datasets, i.e., 10 subjects collected at their institution and 797 3T MR images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Specifically, for the mean Dice ratio of all 10 subjects, the proposed method achieved 94.52% ± 0.9%, 89.49% ± 1.83%, and 79.97% ± 4.32% for WM, GM, and CSF, respectively, which are significantly better than the state-of-the-art methods (p-values < 0.021). For the ADNI dataset, the group difference comparisons indicate that the proposed algorithm outperforms state-of-the-art segmentation methods. CONCLUSIONS: The authors have developed and validated a novel fully automated method for 3T brain MR image segmentation.

Learning-Based 3T Brain MRI Segmentation with Guidance from 7T MRI Labeling.

Brain magnetic resonance image segmentation is one of the most important tasks in medical image analysis and has considerable importance to the effective use of medical imagery in clinical and surgical setting. In particular, the tissue segmentation of white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is crucial for brain measurement and disease diagnosis. A variety of studies have shown that the learning-based techniques are efficient and effective in brain tissue segmentation. However, the learning-based segmentation methods depend largely on the availability of good training labels. The commonly used 3T magnetic resonance (MR) images have insufficient image quality and often exhibit poor intensity contrast between WM, GM, and CSF, therefore not able to provide good training labels for learning-based methods. The advances in ultra-high field 7T imaging make it possible to acquire images with an increasingly high level of quality. In this study, we propose an algorithm based on random forest for segmenting 3T MR images by introducing the segmentation information from their corresponding 7T MR images (through semi-automatic labeling). Furthermore, our algorithm iteratively refines the probability maps of WM, GM, and CSF via a cascade of random forest classifiers to improve the tissue segmentation. Experimental results on 10 subjects with both 3T and 7T MR images in a leave-one-out validation, show that the proposed algorithm performs much better than the state-of-the-art segmentation methods.

Learning-based 3T brain MRI segmentation with guidance from 7T MRI labeling.

PURPOSE: Segmentation of brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is crucial for brain structural measurement and disease diagnosis. Learning-based segmentation methods depend largely on the availability of good training ground truth. However, the commonly used 3T MR images are of insufficient image quality and often exhibit poor intensity contrast between WM, GM, and CSF. Therefore, they are not ideal for providing good ground truth label data for training learning-based methods. Recent advances in ultrahigh field 7T imaging make it possible to acquire images with excellent intensity contrast and signal-to-noise ratio. METHODS: In this paper, the authors propose an algorithm based on random forest for segmenting 3T MR images by training a series of classifiers based on reliable labels obtained semiautomatically from 7T MR images. The proposed algorithm iteratively refines the probability maps of WM, GM, and CSF via a cascade of random forest classifiers for improved tissue segmentation. RESULTS: The proposed method was validated on two datasets, i.e., 10 subjects collected at their institution and 797 3T MR images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Specifically, for the mean Dice ratio of all 10 subjects, the proposed method achieved 94.52% ± 0.9%, 89.49% ± 1.83%, and 79.97% ± 4.32% for WM, GM, and CSF, respectively, which are significantly better than the state-of-the-art methods (p-values < 0.021). For the ADNI dataset, the group difference comparisons indicate that the proposed algorithm outperforms state-of-the-art segmentation methods. CONCLUSIONS: The authors have developed and validated a novel fully automated method for 3T brain MR image segmentation.