Integration and Segregation of Dynamic Functional Connectivity States for Mild Cognitive Impairment Revealed by Graph Theory Indicators

Alterations of Functional Connectivity Dynamics in Affective and Psychotic Disorders

BACKGROUND

Patients with psychosis and patients with depression exhibit widespread neurobiological abnormalities. The analysis of dynamic functional connectivity (dFC) allows for the detection of changes in complex brain activity patterns, providing insights into common and unique processes underlying these disorders.

METHODS

We report the analysis of dFC in a large sample including 127 patients at clinical high risk for psychosis, 142 patients with recent-onset psychosis, 134 patients with recent-onset depression, and 256 healthy control participants. A sliding window-based technique was used to calculate the time-dependent FC in resting-state magnetic resonance imaging data, followed by clustering to reveal recurrent FC states in each diagnostic group.

RESULTS

We identified 5 unique FC states, which could be identified in all groups with high consistency (mean r = 0.889 [SD = 0.116]). Analysis of dynamic parameters of these states showed a characteristic increase in the lifetime and frequency of a weakly connected FC state in patients with recent-onset depression (p < .0005) compared with the other groups and a common increase in the lifetime of an FC state characterized by high sensorimotor and cingulo-opercular connectivities in all patient groups compared with the healthy control group (p < .0002). Canonical correlation analysis revealed a mode that exhibited significant correlations between dFC parameters and clinical variables (r = 0.617, p < .0029), which was associated with positive psychosis symptom severity and several dFC parameters.

CONCLUSIONS

Our findings indicate diagnosis-specific alterations of dFC and underline the potential of dynamic analysis to characterize disorders such as depression and psychosis and clinical risk states.

Altered static and dynamic functional network connectivity in primary angle-closure glaucoma patients

To explore altered patterns of static and dynamic functional brain network connectivity (sFNC and dFNC) in Primary angle-closure glaucoma (PACG) patients. Clinically confirmed 34 PACG patients and 33 age- and gender-matched healthy controls (HCs) underwent evaluation using T1 anatomical and functional MRI on a 3 T scanner. Independent component analysis, sliding window, and the K-means clustering method were employed to investigate the functional network connectivity (FNC) and temporal metrics based on eight resting-state networks. Differences in FNC and temporal metrics were identified and subsequently correlated with clinical variables. For sFNC, compared with HCs, PACG patients showed three decreased interactions, including SMN-AN, SMN-VN and VN-AN pairs. For dFNC, we derived four highly structured states of FC that occurred repeatedly between individual scans and subjects, and the results are highly congruent with sFNC. In addition, PACG patients had a decreased fraction of time in state 3 and negatively correlated with IOP (p < 0.05). PACG patients exhibit abnormalities in both sFNC and dFNC. The high degree of overlap between static and dynamic results suggests the stability of functional connectivity networks in PACG patients, which provide a new perspective to understand the neuropathological mechanisms of optic nerve damage in PACG patients.

Spatial-Temporal Characteristics of Brain Activity in Autism Spectrum Disorder Based on Hidden Markov Model and Dynamic Graph Theory: A Resting-State fMRI Study

Autism spectrum disorder (ASD) is a common neurodevelopmental disorder. Functional magnetic resonance imaging (fMRI) can be used to measure the temporal correlation of blood-oxygen-level-dependent (BOLD) signals in the brain to assess the brain's intrinsic connectivity and capture dynamic changes in the brain. In this study, the hidden Markov model (HMM) and dynamic graph (DG) theory are used to study the spatial-temporal characteristics and dynamics of brain networks based on dynamic functional connectivity (DFC). By using HMM, we identified three typical brain states for ASD and healthy control (HC). Furthermore, we explored the correlation between HMM time-varying properties and clinical autism scale scores. Differences in brain topological characteristics and dynamics between ASD and HC were compared by DG analysis. The experimental results indicate that ASD is more inclined to enter a strongly connected HMM brain state, leading to the isolation of brain networks and alterations in the topological characteristics of brain networks, such as default mode network (DMN), ventral attention network (VAN), and visual network (VN). This work suggests that using different data-driven methods based on DFC to study brain network dynamics would have better information complementarity, which can provide a new direction for the extraction of neuro-biomarkers in the early diagnosis of ASD.

Changes in Community Structure of Brain Dynamic Functional Connectivity States in Mild Cognitive Impairment

Recent researches have noted many changes of short-term dynamic modalities in mild cognitive impairment (MCI) patients' brain functional networks. In this study, the dynamic functional brain networks of 82 MCI patients and 85 individuals in the normal control (NC) group were constructed using the sliding window method and Pearson correlation. The window size was determined using single-scale time-dependent (SSTD) method. Subsequently, k-means was applied to cluster all window samples, identifying three dynamic functional connectivity (DFC) states. Collective sparse symmetric non-negative matrix factorization (cssNMF) was then used to perform community detection on these states and quantify differences in brain regions. Finally, metrics such as within-community connectivity strength, community strength, and node diversity were calculated for further analysis. The results indicated high similarity between the two groups in state 2, with no significant differences in optimal community quantity and functional segregation (p < 0.05). However, for state 1 and state 3, the optimal community quantity was smaller in MCI patients compared to the NC group. In state 1, MCI patients had lower within-community connectivity strength and overall strength than the NC group, whereas state 3 showed results opposite to state 1. Brain regions with statistical difference included MFG.L, ORBinf.R, STG.R, IFGtriang.L, CUN.L, CUN.R, LING.R, SOG.L, and PCUN.R. This study on DFC states explores changes in the brain functional networks of patients with MCI from the perspective of alterations in the community structures of DFC states. The findings could provide new insights into the pathological changes in the brains of MCI patients.

Brain network classification based on dynamic graph attention information bottleneck

BACKGROUND AND OBJECTIVES

Graph neural networks (GNN) have demonstrated remarkable encoding capabilities in the context of brain network classification tasks. They excel at uncovering hidden static connections between brain states. However, brain network signals can be influenced by physiological traits and external variables during clinical detection, resulting in noisy brain graphs. Additionally, many existing algorithms for brain networks primarily focus on static topologies determined by threshold-based criteria, thereby overlooking the real-time variability in brain channel connectivity. These sources of noise and the persistence of static structures inevitably hinder the effective exchange of information during brain network computations.

METHODS

To address these challenges, we propose a novel framework called the dynamic graph attention information bottleneck (DGAIB). This framework is designed to dynamically enhance the input raw brain graph structure from the perspective of information theory and graph theory. First, we employ the Spearman function to construct a raw graph. Then, we use a graph information bottleneck (GIB) to optimize the internal graph connections by selectively masking redundant feature embeddings. Finally, we enhance the feature aggregation of each brain state by utilizing a graph attention network (GAT), which promotes improved information exchange among distinct brain regions within the model. These processed representations serve as input for subsequent classification tasks.

EXPERIMENT AND RESULTS

We systematically evaluated the robustness and generalizability of our proposed framework through a series of experiments. This evaluation included patient-specific experiments using the electroencephalography (EEG)-based CHB-MIT dataset and cross-patient experiments leveraging the functional magnetic resonance imaging (fMRI)-based ABIDE-I dataset from multiple perspectives.

Effect of acupuncture on brain regions modulation of mild cognitive impairment: A meta-analysis of functional magnetic resonance imaging studies

Background

As a non-pharmacological therapy, acupuncture has significant efficacy in treating Mild Cognitive Impairment (MCI) compared to pharmacological therapies. In recent years, advances in neuroimaging techniques have provided new perspectives to elucidate the central mechanisms of acupuncture for MCI. Many acupuncture brain imaging studies have found significant improvements in brain function after acupuncture treatment of MCI, but the underlying mechanisms of brain regions modulation are unclear.

Objective

A meta-analysis of functional magnetic resonance imaging studies of MCI patients treated with acupuncture was conducted to summarize the effects of acupuncture on the modulation of MCI brain regions from a neuroimaging perspective.

Methods

Using acupuncture, neuroimaging, magnetic resonance, and Mild Cognitive Impairment as search terms, PubMed, EMBASE, Web of Science, Cochrane Library, Cochrane Database of Systematic Reviews, Cochrane Database of Abstracts of Reviews of Effects (DARE), Google Scholar, China National Knowledge Infrastructure (CNKI), China Biology Medicine disk (CBM disk), Wanfang and Chinese Scientific Journal Database (VIP) for brain imaging studies on acupuncture on MCI published up to April 2022. Voxel-based neuroimaging meta-analysis of fMRI data was performed using voxel-based d Mapping with Permutation of Subject Images (SDM-PSI), allowing for Family-Wise Error Rate (FWER) correction correction for correction multiple comparisons of results. Subgroup analysis was used to compare the differences in brain regions between the acupuncture treatment group and other control groups. Meta-regression was used to explore demographic information and altered cognitive function effects on brain imaging outcomes. Linear models were drawn using MATLAB 2017a, and visual graphs for quality evaluation were produced using R software and RStudio software.

Results

A total of seven studies met the inclusion criteria, with 94 patients in the treatment group and 112 patients in the control group. All studies were analyzed using the regional homogeneity (ReHo) method. The experimental design of fMRI included six task state studies and one resting-state study. The meta-analysis showed that MCI patients had enhanced activity in the right insula, left anterior cingulate/paracingulate gyri, right thalamus, right middle frontal gyrus, right median cingulate/paracingulate gyri, and right middle temporal gyrus brain regions after acupuncture treatment. Further analysis of RCT and longitudinal studies showed that Reho values were significantly elevated in two brain regions, the left anterior cingulate/paracingulate gyrus and the right insula, after acupuncture. The MCI group showed stronger activity in the right supramarginal gyrus after acupuncture treatment compared to healthy controls. Meta-regression analysis showed that the right anterior thalamic projection ReHo index was significantly correlated with the MMSE score after acupuncture treatment in all MCI patients.

Conclusions

Acupuncture therapy has a modulating effect on the brain regions of MCI patients. However, due to the inadequate experimental design of neuroimaging studies, multi-center neuroimaging studies with large samples are needed better to understand the potential neuroimaging mechanisms of acupuncture for MCI. In addition, machine learning algorithm-based predictive models for evaluating the efficacy of acupuncture for MCI may become a focus of future research.

Systematic review registration

https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022287826, identifier: CRD 42022287826.

Suboptimal states and frontoparietal network-centered incomplete compensation revealed by dynamic functional network connectivity in patients with post-stroke cognitive impairment

Background

Neural reorganization occurs after a stroke, and dynamic functional network connectivity (dFNC) pattern is associated with cognition. We hypothesized that dFNC alterations resulted from neural reorganization in post-stroke cognitive impairment (PSCI) patients, and specific dFNC patterns characterized different pathological types of PSCI.

Methods

Resting-state fMRI data were collected from 16 PSCI patients with hemorrhagic stroke (hPSCI group), 21 PSCI patients with ischemic stroke (iPSCI group), and 21 healthy controls (HC). We performed the dFNC analysis for the dynamic connectivity states, together with their topological and temporal features.

Results

We identified 10 resting-state networks (RSNs), and the dFNCs could be clustered into four reoccurring states (modular, regional, sparse, and strong). Compared with HC, the hPSCI and iPSCI patients showed lower standard deviation (SD) and coefficient of variation (CV) in the regional and modular states, respectively (p &lt; 0.05). Reduced connectivities within the primary network (visual, auditory, and sensorimotor networks) and between the primary and high-order cognitive control domains were observed (p &lt; 0.01).

Conclusion

The transition trend to suboptimal states may play a compensatory role in patients with PSCI through redundancy networks. The reduced exploratory capacity (SD and CV) in different suboptimal states characterized cognitive impairment and pathological types of PSCI. The functional disconnection between the primary and high-order cognitive control network and the frontoparietal network centered (FPN-centered) incomplete compensation may be the pathological mechanism of PSCI. These results emphasize the flexibility of neural reorganization during self-repair.

Research on Voxel-Based Features Detection and Analysis of Alzheimer’s Disease Using Random Survey Support Vector Machine

Alzheimer’s disease (AD) is a degenerative disease of the central nervous system characterized by memory and cognitive dysfunction, as well as abnormal changes in behavior and personality. The research focused on how machine learning classified AD became a recent hotspot. In this study, we proposed a novel voxel-based feature detection framework for AD. Specifically, using 649 voxel-based morphometry (VBM) methods obtained from MRI in Alzheimer’s Disease Neuroimaging Initiative (ADNI), we proposed a feature detection method according to the Random Survey Support Vector Machines (RS-SVM) and combined the research process based on image-, gene-, and pathway-level analysis for AD prediction. Particularly, we constructed 136, 141, and 113 novel voxel-based features for EMCI (early mild cognitive impairment)-HC (healthy control), LMCI (late mild cognitive impairment)-HC, and AD-HC groups, respectively. We applied linear regression model, least absolute shrinkage and selection operator (Lasso), partial least squares (PLS), SVM, and RS-SVM five methods to test and compare the accuracy of these features in these three groups. The prediction accuracy of the AD-HC group using the RS-SVM method was higher than 90%. In addition, we performed functional analysis of the features to explain the biological significance. The experimental results using five machine learning indicate that the identified features are effective for AD and HC classification, the RS-SVM framework has the best classification accuracy, and our strategy can identify important brain regions for AD.

BCNet: A Novel Network for Blood Cell Classification

Aims: Most blood diseases, such as chronic anemia, leukemia (commonly known as blood cancer), and hematopoietic dysfunction, are caused by environmental pollution, substandard decoration materials, radiation exposure, and long-term use certain drugs. Thus, it is imperative to classify the blood cell images. Most cell classification is based on the manual feature, machine learning classifier or the deep convolution network neural model. However, manual feature extraction is a very tedious process, and the results are usually unsatisfactory. On the other hand, the deep convolution neural network is usually composed of massive layers, and each layer has many parameters. Therefore, each deep convolution neural network needs a lot of time to get the results. Another problem is that medical data sets are relatively small, which may lead to overfitting problems. Methods: To address these problems, we propose seven models for the automatic classification of blood cells: BCARENet, BCR5RENet, BCMV2RENet, BCRRNet, BCRENet, BCRSNet, and BCNet. The BCNet model is the best model among the seven proposed models. The backbone model in our method is selected as the ResNet-18, which is pre-trained on the ImageNet set. To improve the performance of the proposed model, we replace the last four layers of the trained transferred ResNet-18 model with the three randomized neural networks (RNNs), which are RVFL, ELM, and SNN. The final outputs of our BCNet are generated by the ensemble of the predictions from the three randomized neural networks by the majority voting. We use four multi-classification indexes for the evaluation of our model. Results: The accuracy, average precision, average F1-score, and average recall are 96.78, 97.07, 96.78, and 96.77%, respectively. Conclusion: We offer the comparison of our model with state-of-the-art methods. The results of the proposed BCNet model are much better than other state-of-the-art methods.

Research on Pathogenic Hippocampal Voxel Detection in Alzheimer's Disease Using Clustering Genetic Random Forest

Alzheimer's disease (AD) is an age-related neurological disease, which is closely associated with hippocampus, and subdividing the hippocampus into voxels can capture subtle signals that are easily missed by region of interest (ROI) methods. Therefore, studying interpretable associations between voxels can better understand the effect of voxel set on the hippocampus and AD. In this study, by analyzing the hippocampal voxel data, we propose a novel method based on clustering genetic random forest to identify the important voxels. Specifically, we divide the left and right hippocampus into voxels to constitute the initial feature set. Moreover, the random forest is constructed using the randomly selected samples and features. The genetic evolution is used to amplify the difference in decision trees and the clustering evolution is applied to generate offspring in genetic evolution. The important voxels are the features that reach the peak classification. The results demonstrate that our method has good classification and stability. Particularly, through biological analysis of the obtained voxel set, we find that they play an important role in AD by affecting the function of the hippocampus. These discoveries demonstrate the contribution of the voxel set to AD.

A Triple-Network Dynamic Connection Study in Alzheimer's Disease

Alzheimer's disease (AD) was associated with abnormal organization and function of large-scale brain networks. We applied group independent component analysis (Group ICA) to construct the triple-network consisting of the saliency network (SN), the central executive network (CEN), and the default mode network (DMN) in 25 AD, 60 mild cognitive impairment (MCI) and 60 cognitively normal (CN) subjects. To explore the dynamic functional network connectivity (dFNC), we investigated dynamic time-varying triple-network interactions in subjects using Group ICA analysis based on k-means clustering (GDA-k-means). The mean of brain state-specific network interaction indices (meanNII) in the three groups (AD, MCI, CN) showed significant differences by ANOVA analysis. To verify the robustness of the findings, a support vector machine (SVM) was taken meanNII, gender and age as features to classify. This method obtained accuracy values of 95, 94, and 77% when classifying AD vs. CN, AD vs. MCI, and MCI vs. CN, respectively. In our work, the findings demonstrated that the dynamic characteristics of functional interactions of the triple-networks contributed to studying the underlying pathophysiology of AD. It provided strong evidence for dysregulation of brain dynamics of AD.