Feature Selection and Combination of Information in the Functional Brain Connectome for Discrimination of Mild Cognitive Impairment and Analyses of Altered Brain Patterns

The White Matter Integrity and Functional Connection Differences of Fornix (Cres)/Stria Terminalis in Individuals with Mild Cognitive Impairment Induced by Occupational Aluminum Exposure

Long-term aluminum (Al) exposure increases the risk of mild cognitive impairment (MCI). The aim of the present study was to investigate the neural mechanisms of Al-induced MCI. In our study, a total of 52 individuals with occupational Al exposure >10 years were enrolled and divided into two groups: MCI (Al-MCI) and healthy controls (Al-HC). Plasma Al concentrations and Montreal Cognitive Assessment (MoCA) score were collected for all participants. And diffusion tensor imaging and resting-state functional magnetic resonance imaging were used to examine changes of white matter (WM) and functional connectivity (FC). There was a negative correlation between MoCA score and plasma Al concentration. Compared with the Al-HC, fractional anisotropy value for the right fornix (cres)/stria terminalis (FX/ST) was higher in the Al-MCI. Furthermore, there was a difference in FC between participants with and without MCI under Al exposure. We defined the regions with differing FC as a "pathway," specifically the connectivity from the right temporal pole to the right FX/ST, then to the right sagittal stratum, and further to the right anterior cingulate and paracingulate gyri and right inferior frontal gyrus, orbital part. In summary, we believe that the observed differences in WM integrity and FC in the right FX/ST between participants with and without MCI under long-term Al exposure may represent the neural mechanisms underlying MCI induced by Al exposure.

A Learnable Counter-Condition Analysis Framework for Functional Connectivity-Based Neurological Disorder Diagnosis

To understand the biological characteristics of neurological disorders with functional connectivity (FC), recent studies have widely utilized deep learning-based models to identify the disease and conducted post-hoc analyses via explainable models to discover disease-related biomarkers. Most existing frameworks consist of three stages, namely, feature selection, feature extraction for classification, and analysis, where each stage is implemented separately. However, if the results at each stage lack reliability, it can cause misdiagnosis and incorrect analysis in afterward stages. In this study, we propose a novel unified framework that systemically integrates diagnoses (i.e., feature selection and feature extraction) and explanations. Notably, we devised an adaptive attention network as a feature selection approach to identify individual-specific disease-related connections. We also propose a functional network relational encoder that summarizes the global topological properties of FC by learning the inter-network relations without pre-defined edges between functional networks. Last but not least, our framework provides a novel explanatory power for neuroscientific interpretation, also termed counter-condition analysis. We simulated the FC that reverses the diagnostic information (i.e., counter-condition FC): converting a normal brain to be abnormal and vice versa. We validated the effectiveness of our framework by using two large resting-state functional magnetic resonance imaging (fMRI) datasets, Autism Brain Imaging Data Exchange (ABIDE) and REST-meta-MDD, and demonstrated that our framework outperforms other competing methods for disease identification. Furthermore, we analyzed the disease-related neurological patterns based on counter-condition analysis.

Topological properties analysis and identification of mild cognitive impairment based on individual morphological brain network connectome

Mild cognitive impairment is considered the prodromal stage of Alzheimer's disease. Accurate diagnosis and the exploration of the pathological mechanism of mild cognitive impairment are extremely valuable for targeted Alzheimer's disease prevention and early intervention. In all, 100 mild cognitive impairment patients and 86 normal controls were recruited in this study. We innovatively constructed the individual morphological brain networks and derived multiple brain connectome features based on 3D-T1 structural magnetic resonance imaging with the Jensen-Shannon divergence similarity estimation method. Our results showed that the most distinguishing morphological brain connectome features in mild cognitive impairment patients were consensus connections and nodal graph metrics, mainly located in the frontal, occipital, limbic lobes, and subcortical gray matter nuclei, corresponding to the default mode network. Topological properties analysis revealed that mild cognitive impairment patients exhibited compensatory changes in the frontal lobe, while abnormal cortical-subcortical circuits associated with cognition were present. Moreover, the combination of multidimensional brain connectome features using multiple kernel-support vector machine achieved the best classification performance in distinguishing mild cognitive impairment patients and normal controls, with an accuracy of 84.21%. Therefore, our findings are of significant importance for developing potential brain imaging biomarkers for early detection of Alzheimer's disease and understanding the neuroimaging mechanisms of the disease.

Wavelet transform-based frequency self-adaptive model for functional brain network

The accurate estimation of functional brain networks is essential for comprehending the intricate relationships between different brain regions. Conventional methods such as Pearson Correlation and Sparse Representation often fail to uncover concealed information within diverse frequency bands. To address this limitation, we introduce a novel frequency-adaptive model based on wavelet transform, enabling selective capture of highly correlated frequency band sequences. Our approach involves decomposing the original time-domain signal from resting-state functional magnetic resonance imaging into distinct frequency domains, thus constructing an adjacency matrix that offers enhanced separation of features across brain regions. Comparative analysis demonstrates the superior performance of our proposed model over conventional techniques, showcasing improved clarity and distinctiveness. Notably, we achieved the highest accuracy rate of 89.01% using Sparse Representation based on Wavelet Transform, outperforming Pearson Correlation based on Wavelet Transform with an accuracy of 81.32%. Importantly, our method optimizes raw data without significantly altering feature topology, rendering it adaptable to various functional brain network estimation approaches. Overall, this innovation holds the potential to advance the understanding of brain function and furnish more accurate samples for future research and clinical applications.

Motor intent recognition of multi-feature fusion EEG signals by UMAP algorithm

The key to the analysis of electroencephalogram (EEG) signals lies in the extraction of effective features from the raw EEG signals, which can then be utilized to augment the classification accuracy of motor imagery (MI) applications in brain-computer interface (BCI). It can be argued that the utilization of features from multiple domains can be a more effective approach to feature extraction for MI pattern classification, as it can provide a more comprehensive set of information that the traditional single feature extraction method may not be able to capture. In this paper, a multi-feature fusion algorithm based on uniform manifold approximate and projection (UMAP) is proposed for motor imagery EEG signals. The brain functional network and common spatial pattern (CSP) are initially extracted as features. Subsequently, UMAP is utilized to fuse the extracted multi-domain features to generate low-dimensional features with improved discriminative capability. Finally, the k-nearest neighbor (KNN) classifier is applied in a lower dimensional space. The proposed method is evaluated using left-right hand EEG signals, and achieved the average accuracy of over 92%. The results indicate that, compared with single-domain-based feature extraction methods, multi-feature fusion EEG signal classification based on the UMAP algorithm yields superior classification and visualization performance. Feature extraction and fusion based on UMAP algorithm of left-right hand motor imagery.

Effects of computerized cognitive training on structure‒function coupling and topology of multiple brain networks in people with mild cognitive impairment: a randomized controlled trial

BACKGROUND

People with mild cognitive impairment (MCI) experience a loss of cognitive functions, whose mechanism is characterized by aberrant structure‒function (SC-FC) coupling and topological attributes of multiple networks. This study aimed to reveal the network-level SC-FC coupling and internal topological changes triggered by computerized cognitive training (CCT) to explain the therapeutic effects of this training in individuals with MCI.

METHODS

In this randomized block experiment, we recruited 60 MCI individuals and randomly divided them into an 8-week multidomain CCT group and a health education control group. The neuropsychological outcome measures were the Montreal Cognitive Assessment (MoCA), Chinese Auditory Verbal Learning Test (CAVLT), Chinese Stroop Color-Word Test (SCWT), and Rey-Osterrieth Complex Figure Test (Rey CFT). The brain imaging outcome measures were SC-FC coupling and topological attributes using functional MRI and diffusion tensor imaging methods. We applied linear model analysis to assess the differences in the outcome measures and identify the correspondence between the changes in the brain networks and cognitive functions before and after the CCT.

RESULTS

Fifty participants were included in the analyses after the exclusion of three dropouts and seven participants with low-quality MRI scans. Significant group × time effects were found on the changes in the MoCA, CAVLT, and Rey CFT recall scores. The changes in the SC-FC coupling values of the default mode network (DMN) and somatomotor network (SOM) were higher in the CCT group than in the control group (P(unc.) = 0.033, P(unc.) = 0.019), but opposite effects were found on the coupling values of the visual network (VIS) (P(unc.) = 0.039). Increasing clustering coefficients in the functional DMN and SOM and subtle changes in the nodal degree centrality and nodal efficiency of the right dorsal medial prefrontal cortex, posterior cingulate cortex, left parietal lobe, somatomotor area, and visual cortex were observed in the CCT group (P < 0.05, Bonferroni correction). Significant correspondences were found between global cognitive function and DMN coupling values (P(unc.) = 0.007), between immediate memory and SOM as well as FPC coupling values (P(unc.) = 0.037, P(unc.) = 0.030), between delayed memory and SOM coupling values (P(unc.) = 0.030), and between visual memory and VIS coupling values (P(unc.) = 0.007).

CONCLUSIONS

Eight weeks of CCT effectively improved global cognitive and memory functions; these changes were correlated with increases in SC-FC coupling and changes in the topography of the DMN and SOM in individuals with MCI. The CCT regimen also modulated the clustering coefficient and the capacity for information transformation in functional networks; these effects appeared to underlie the cognitive improvement associated with CCT.

TRIAL REGISTRATION

Chinese Clinical Trial Registry, ChiCTR2000034012. Registered on 21 June 2020.

Brain disease research based on functional magnetic resonance imaging data and machine learning: a review

Brain diseases, including neurodegenerative diseases and neuropsychiatric diseases, have long plagued the lives of the affected populations and caused a huge burden on public health. Functional magnetic resonance imaging (fMRI) is an excellent neuroimaging technology for measuring brain activity, which provides new insight for clinicians to help diagnose brain diseases. In recent years, machine learning methods have displayed superior performance in diagnosing brain diseases compared to conventional methods, attracting great attention from researchers. This paper reviews the representative research of machine learning methods in brain disease diagnosis based on fMRI data in the recent three years, focusing on the most frequent four active brain disease studies, including Alzheimer's disease/mild cognitive impairment, autism spectrum disorders, schizophrenia, and Parkinson's disease. We summarize these 55 articles from multiple perspectives, including the effect of the size of subjects, extracted features, feature selection methods, classification models, validation methods, and corresponding accuracies. Finally, we analyze these articles and introduce future research directions to provide neuroimaging scientists and researchers in the interdisciplinary fields of computing and medicine with new ideas for AI-aided brain disease diagnosis.

Multimodal Classification Framework Based on Hypergraph Latent Relation for End-Stage Renal Disease Associated with Mild Cognitive Impairment

Combined arterial spin labeling (ASL) and functional magnetic resonance imaging (fMRI) can reveal more comprehensive properties of the spatiotemporal and quantitative properties of brain networks. Imaging markers of end-stage renal disease associated with mild cognitive impairment (ESRDaMCI) will be sought from these properties. The current multimodal classification methods often neglect to collect high-order relationships of brain regions and remove noise from the feature matrix. A multimodal classification framework is proposed to address this issue using hypergraph latent relation (HLR). A brain functional network with hypergraph structural information is constructed by fMRI data. The feature matrix is obtained through graph theory (GT). The cerebral blood flow (CBF) from ASL is selected as the second modal feature matrix. Then, the adaptive similarity matrix is constructed by learning the latent relation between feature matrices. Latent relation adaptive similarity learning (LRAS) is introduced to multi-task feature learning to construct a multimodal feature selection method based on latent relation (LRMFS). The experimental results show that the best classification accuracy (ACC) reaches 88.67%, at least 2.84% better than the state-of-the-art methods. The proposed framework preserves more valuable information between brain regions and reduces noise among feature matrixes. It provides an essential reference value for ESRDaMCI recognition.

The reconfiguration pattern of individual brain metabolic connectome for Parkinson's disease identification

18F-Fluorodeoxyglucose positron emission tomography (18F-FDG PET) is widely employed to reveal metabolic abnormalities linked to Parkinson's disease (PD) at a systemic level. However, the individual metabolic connectome details with PD based on 18F-FDG PET remain largely unknown. To alleviate this issue, we derived a novel brain network estimation method for individual metabolic connectome, that is, Jensen-Shannon Divergence Similarity Estimation (JSSE). Further, intergroup difference between the individual's metabolic brain network and its global/local graph metrics was analyzed to investigate the metabolic connectome's alterations. To further improve the PD diagnosis performance, multiple kernel support vector machine (MKSVM) is conducted for identifying PD from normal control (NC), which combines both topological metrics and connection. Resultantly, PD individuals showed higher nodal topological properties (including assortativity, modularity score, and characteristic path length) than NC individuals, whereas global efficiency and synchronization were lower. Moreover, 45 most significant connections were affected. Further, consensus connections in occipital, parietal, and frontal regions were decrease in PD while increase in subcortical, temporal, and prefrontal regions. The abnormal metabolic network measurements depicted an ideal classification in identifying PD of NC with an accuracy up to 91.84%. The JSSE method identified the individual-level metabolic connectome of 18F-FDG PET, providing more dimensional and systematic mechanism insights for PD.

Predicting sex, age, general cognition and mental health with machine learning on brain structural connectomes

There is an increasing expectation that advanced, computationally expensive machine learning (ML) techniques, when applied to large population-wide neuroimaging datasets, will help to uncover key differences in the human brain in health and disease. We take a comprehensive approach to explore how multiple aspects of brain structural connectivity can predict sex, age, general cognitive function and general psychopathology, testing different ML algorithms from deep learning (DL) model (BrainNetCNN) to classical ML methods. We modelled N = 8183 structural connectomes from UK Biobank using six different structural network weightings obtained from diffusion MRI. Streamline count generally provided the highest prediction accuracies in all prediction tasks. DL did not improve on prediction accuracies from simpler linear models. Further, high correlations between gradient attribution coefficients from DL and model coefficients from linear models suggested the models ranked the importance of features in similar ways, which indirectly suggested the similarity in models' strategies for making predictive decision to some extent. This highlights that model complexity is unlikely to improve detection of associations between structural connectomes and complex phenotypes with the current sample size.

Increased functional connectivity patterns in mild Alzheimer's disease: A rsfMRI study

Background

Alzheimer's disease (AD) is the most common age-related neurodegenerative disorder. In view of our rapidly aging population, there is an urgent need to identify Alzheimer's disease (AD) at an early stage. A potential way to do so is by assessing the functional connectivity (FC), i.e., the statistical dependency between two or more brain regions, through novel analysis techniques.

Methods

In the present study, we assessed the static and dynamic FC using different approaches. A resting state (rs)fMRI dataset from the Alzheimer's disease neuroimaging initiative (ADNI) was used (n = 128). The blood-oxygen-level-dependent (BOLD) signals from 116 regions of 4 groups of participants, i.e., healthy controls (HC; n = 35), early mild cognitive impairment (EMCI; n = 29), late mild cognitive impairment (LMCI; n = 30), and Alzheimer's disease (AD; n = 34) were extracted and analyzed. FC and dynamic FC were extracted using Pearson's correlation, sliding-windows correlation analysis (SWA), and the point process analysis (PPA). Additionally, graph theory measures to explore network segregation and integration were computed.

Results

Our results showed a longer characteristic path length and a decreased degree of EMCI in comparison to the other groups. Additionally, an increased FC in several regions in LMCI and AD in contrast to HC and EMCI was detected. These results suggest a maladaptive short-term mechanism to maintain cognition.

Conclusion

The increased pattern of FC in several regions in LMCI and AD is observable in all the analyses; however, the PPA enabled us to reduce the computational demands and offered new specific dynamic FC findings.

Morphologic brain network predicts levodopa responsiveness in Parkinson disease

Background

The levodopa challenge test (LCT) has been routinely used in Parkinson disease (PD) evaluation and predicts the outcome of deep brain stimulation (DBS). Guidelines recommend that patients with an improvement in Unified Parkinson's Disease Rating Scale (UPDRS)-III score > 33% in the LCT receive DBS treatment. However, LCT results are affected by many factors, and only provide information on the immediate effectiveness of dopamine. The aim of the present study was to investigate the relationship between LCT outcome and brain imaging features of PD patients to determine whether the latter can be used to identify candidates for DBS.

Methods

A total of 38 PD patients were enrolled in the study. Based on improvement in UPDRS-III score in the LCT, patients were divided into low improvement (PD-LCT-L) and high improvement (PD-LCT-H) groups. Each patient's neural network was reconstructed based on T1-weighted magnetic resonance imaging data using the Jensen-Shannon divergence similarity estimation method. The network was established with the multiple kernel support vector machine technique. We analyzed differences in individual morphologic brain networks and their global and local metrics to determine whether there were differences in the connectomes of PD-LCT-L and PD-LCT-H groups.

Results

The 2 groups were similar in terms of demographic and clinical characteristics. Mean ± SD levodopa responsiveness was 26.52% ± 3.47% in the PD-LCT-L group (N = 13) and 58.66% ± 4.09% in the PD-LCT-H group (N = 25). There were no significant differences between groups in global and local metrics. There were 43 consensus connections that were affected in both groups; in PD-LCT-L patients, most of these connections were decreased whereas those related to the dorsolateral superior frontal gyrus and left cuneus were significantly increased.

Conclusion

Morphologic brain network assessment is a valuable method for predicting levodopa responsiveness in PD patients, which can facilitate the selection of candidates for DBS.

Classification and Interpretability of Mild Cognitive Impairment Based on Resting-State Functional Magnetic Resonance and Ensemble Learning

The combination and integration of multimodal imaging and clinical markers have introduced numerous classifiers to improve diagnostic accuracy in detecting and predicting AD; however, many studies cannot ensure the homogeneity of data sets and consistency of results. In our study, the XGBoost algorithm was used to classify mild cognitive impairment (MCI) and normal control (NC) populations through five rs-fMRI analysis datasets. Shapley Additive exPlanations (SHAP) is used to analyze the interpretability of the model. The highest accuracy for diagnosing MCI was 65.14% (using the mPerAF dataset). The characteristics of the left insula, right middle frontal gyrus, and right cuneus correlated positively with the output value using DC datasets. The characteristics of left cerebellum 6, right inferior frontal gyrus, opercular part, and vermis 6 correlated positively with the output value using fALFF datasets. The characteristics of the right middle temporal gyrus, left middle temporal gyrus, left temporal pole, and middle temporal gyrus correlated positively with the output value using mPerAF datasets. The characteristics of the right middle temporal gyrus, left middle temporal gyrus, and left hippocampus correlated positively with the output value using PerAF datasets. The characteristics of left cerebellum 9, vermis 9, and right precentral gyrus, right amygdala, and left middle occipital gyrus correlated positively with the output value using Wavelet-ALFF datasets. We found that the XGBoost algorithm constructed from rs-fMRI data is effective for the diagnosis and classification of MCI. The accuracy rates obtained by different rs-fMRI data analysis methods are similar, but the important features are different and involve multiple brain regions, which suggests that MCI may have a negative impact on brain function.

Altered pattern analysis and identification of subjective cognitive decline based on morphological brain network

Subjective cognitive decline (SCD) is considered the first stage of Alzheimer's disease (AD). Accurate diagnosis and the exploration of the pathological mechanism of SCD are extremely valuable for targeted AD prevention. However, there is little knowledge of the specific altered morphological network patterns in SCD individuals. In this present study, 36 SCD cases and 34 paired-matched normal controls (NCs) were recruited. The Jensen-Shannon distance-based similarity (JSS) method was implemented to construct and derive the attributes of multiple brain connectomes (i.e., morphological brain connections and global and nodal graph metrics) of individual morphological brain networks. A t-test was used to discriminate between the selected nodal graph metrics, while the leave-one-out cross-validation (LOOCV) was used to obtain consensus connections. Comparisons were performed to explore the altered patterns of connectome features. Further, the multiple kernel support vector machine (MK-SVM) was used for combining brain connectomes and differentiating SCD from NCs. We showed that the consensus connections and nodal graph metrics with the most discriminative ability were mostly found in the frontal, limbic, and parietal lobes, corresponding to the default mode network (DMN) and frontoparietal task control (FTC) network. Altered pattern analysis demonstrated that SCD cases had a tendency for modularity and local efficiency enhancement. Additionally, using the MK-SVM to combine the features of multiple brain connectomes was associated with optimal classification performance [area under the curve (AUC): 0.9510, sensitivity: 97.22%, specificity: 85.29%, and accuracy: 91.43%]. Therefore, our study highlighted the combination of multiple connectome attributes based on morphological brain networks and offered a valuable method for distinguishing SCD individuals from NCs. Moreover, the altered patterns of multidimensional connectome attributes provided a promising insight into the neuroimaging mechanism and early intervention in SCD subjects.

Correlation Between Serum High-Sensitivity C-Reactive Protein, Tumor Necrosis Factor-Alpha, Serum Interleukin-6 and White Matter Integrity Before and After the Treatment of Drug-Naïve Patients With Major Depressive Disorder

Background

Previous studies have noticed that systemic inflammation may alter the integrity of white matter. However, how the levels of serum cytokine affect the integrity of white matter in major depressive disorder (MDD) patients are unclear. Our study aimed to investigate the association between the inflammatory cytokine levels and white matter microstructure in drug-naïve patients with MDD pre- and post-treatment.

Method

In total, 29 MDD patients and 25 healthy controls (HC) were included in this study. Diffusion tensor imaging (DTI) was conducted in all subjects at baseline, and the MDD patients were reassessed after venlafaxine treatment, using a tract-based spatial statistics (TBSS) analysis. Morning serum interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α), and high-sensitivity C-reactive protein (hs-CRP) concentrations in MDD patients were also measured pre- and post-treatment.

Results

Significantly reduced fractional anisotropy (FA) values were found in the bilateral superior fronto-occipital fasciculus (SFO), posterior limb of the internal capsule (IC-PL), and fornix compared with the HC, and FA values in these regions in MDD patients have risen to normal levels except the bilateral SFO after treatment. The FA value of the left IC-PL was inversely correlated with the peripheral hs-CRP levels in both pre- and post-treatment MDD patients.

Conclusion

Our results suggested that the white matter integrity in the left IC-PL was significantly inversely correlated with the peripheral hs-CRP levels in both pre- and post-treatment MDD patients.

Individual Brain Morphological Connectome Indicator Based on Jensen-Shannon Divergence Similarity Estimation for Autism Spectrum Disorder Identification

Background

Structural magnetic resonance imaging (sMRI) reveals abnormalities in patients with autism spectrum syndrome (ASD). Previous connectome studies of ASD have failed to identify the individual neuroanatomical details in preschool-age individuals. This paper aims to establish an individual morphological connectome method to characterize the connectivity patterns and topological alterations of the individual-level brain connectome and their diagnostic value in patients with ASD.

Methods

Brain sMRI data from 24 patients with ASD and 17 normal controls (NCs) were collected; participants in both groups were aged 24-47 months. By using the Jensen-Shannon Divergence Similarity Estimation (JSSE) method, all participants's morphological brain network were ascertained. Student's t-tests were used to extract the most significant features in morphological connection values, global graph measurement, and node graph measurement.

Results

The results of global metrics' analysis showed no statistical significance in the difference between two groups. Brain regions with meaningful properties for consensus connections and nodal metric features are mostly distributed in are predominantly distributed in the basal ganglia, thalamus, and cortical regions spanning the frontal, temporal, and parietal lobes. Consensus connectivity results showed an increase in most of the consensus connections in the frontal, parietal, and thalamic regions of patients with ASD, while there was a decrease in consensus connectivity in the occipital, prefrontal lobe, temporal lobe, and pale regions. The model that combined morphological connectivity, global metrics, and node metric features had optimal performance in identifying patients with ASD, with an accuracy rate of 94.59%.

Conclusion

The individual brain network indicator based on the JSSE method is an effective indicator for identifying individual-level brain network abnormalities in patients with ASD. The proposed classification method can contribute to the early clinical diagnosis of ASD.

Individual-specific networks for prediction modelling – A scoping review of methods

Background

Recent advances in biotechnology enable the acquisition of high-dimensional data on individuals, posing challenges for prediction models which traditionally use covariates such as clinical patient characteristics. Alternative forms of covariate representations for the features derived from these modern data modalities should be considered that can utilize their intrinsic interconnection. The connectivity information between these features can be represented as an individual-specific network defined by a set of nodes and edges, the strength of which can vary from individual to individual. Global or local graph-theoretical features describing the network may constitute potential prognostic biomarkers instead of or in addition to traditional covariates and may replace the often unsuccessful search for individual biomarkers in a high-dimensional predictor space.

Methods

We conducted a scoping review to identify, collate and critically appraise the state-of-art in the use of individual-specific networks for prediction modelling in medicine and applied health research, published during 2000–2020 in the electronic databases PubMed, Scopus and Embase.

Results

Our scoping review revealed the main application areas namely neurology and pathopsychology, followed by cancer research, cardiology and pathology (N = 148). Network construction was mainly based on Pearson correlation coefficients of repeated measurements, but also alternative approaches (e.g. partial correlation, visibility graphs) were found. For covariates measured only once per individual, network construction was mostly based on quantifying an individual’s contribution to the overall group-level structure. Despite the multitude of identified methodological approaches for individual-specific network inference, the number of studies that were intended to enable the prediction of clinical outcomes for future individuals was quite limited, and most of the models served as proof of concept that network characteristics can in principle be useful for prediction.

Conclusion

The current body of research clearly demonstrates the value of individual-specific network analysis for prediction modelling, but it has not yet been considered as a general tool outside the current areas of application. More methodological research is still needed on well-founded strategies for network inference, especially on adequate network sparsification and outcome-guided graph-theoretical feature extraction and selection, and on how networks can be exploited efficiently for prediction modelling.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-022-01544-6.

Rich-Club Organization Disturbances of the Individual Morphological Network in Subjective Cognitive Decline

Background

Subjective cognitive decline (SCD) was considered to be the preclinical stage of Alzheimer's disease (AD). However, less is known about the altered rich-club organizations of the morphological networks in individuals with SCD.

Methods

This study included 53 individuals with SCD and 54 well-matched healthy controls (HC) from the Alzheimer's disease Neuroimaging Initiative (ADNI) database. Individual-level brain morphological networks were constructed by estimating the Jensen-Shannon distance-based similarity in the distribution of regional gray matter volume. Rich-club properties were then detected, followed by statistical comparison.

Results

The characteristic rich-club organization of morphological networks (normalized rich-club coefficients > 1) was observed for both the SCD and HC groups under a range of thresholds. The SCD group showed a reduced normalized rich-club coefficient compared with the HC group. The SCD group exhibited the decreased strength and degree of rich-club connections than the HC group (strength: HC = 79.93, SCD = 74.37, p = 0.028; degree: HC = 85.28, SCD = 79.34, p = 0.027). Interestingly, the SCD group showed an increased strength of local connections than the HC group (strength: HC = 1982.16, SCD = 2003.38, p = 0.036).

Conclusion

Rich-club organization disturbances of morphological networks in individuals with SCD reveal a distinct pattern between the rich-club and peripheral regions. This altered rich-club organization pattern provides novel insights into the underlying mechanism of SCD and could be used to investigate prevention strategies at the preclinical stage of AD.

Not Only in Sensorimotor Network: Local and Distant Cerebral Inherent Activity of Chronic Ankle Instability—A Resting-State fMRI Study

Background

Increasing evidence has proved that chronic ankle instability (CAI) is highly related to the central nervous system (CNS). However, it is still unclear about the inherent cerebral activity among the CAI patients.

Purpose

To investigate the differences of intrinsic functional cerebral activity between the CAI patients and healthy controls (HCs) and further explore its correlation with clinical measurement in CAI patients.

Materials and Methods

A total of 25 CAI patients and 39 HCs were enrolled in this study. Resting-state functional magnetic resonance imaging (rs-fMRI) was used to detect spontaneous cerebral activity. The metrics of amplitude of low-frequency fluctuation (ALFF), fractional ALFF (fALFF), and regional homogeneity (ReHo) of the two groups were compared by two-sample t-test. The brain regions that demonstrated altered functional metrics were selected as the regions of interest (ROIs). The functional connectivity (FC) was analyzed based on the ROIs. The Spearman correlation was calculated between rs-fMRI metrics and clinical scale scores.

Results

Compared with HCs, CAI patients showed higher ALFF and ReHo values in the right postcentral gyrus, the right precentral gyrus, and the right middle frontal gyrus, while lower fALFF values in the orbital-frontal cortex (OFC, p &lt; 0.01 after correction). Increasing FC between the right precentral gyrus and the right postcentral gyrus while decreasing FC between the right precentral gyrus and the anterior cingulum cortex (ACC), the right middle frontal gyrus and the left middle temporal gyrus, and the OFC and left inferior parietal lobule (IPL) was observed. In addition, in the CAI group, the ReHo value negatively correlated with the Cumberland Ankle Instability Tool score in the right middle frontal gyrus (r = −0.52, p = 0.007).

Conclusion

The CAI patients exhibited enhanced and more coherent regional inherent neuronal activity within the sensorimotor network while lower regional inherent activity in pain/emotion modulation related region. In addition, the information exchanges were stronger within the sensorimotor network while weaker between distant interhemispheric regions. Besides, the increased inherent activity in the right middle frontal gyrus was related to clinical severity. These findings may provide insights into the pathophysiological alteration in CNS among CAI patients.

A Comparative Study of Structural and Metabolic Brain Networks in Patients With Mild Cognitive Impairment

Background: Changes in the metabolic and structural brain networks in mild cognitive impairment (MCI) have been widely researched. However, few studies have compared the differences in the topological properties of the metabolic and structural brain networks in patients with MCI.

Methods: We analyzedmagnetic resonance imaging (MRI) and fluoro-deoxyglucose positron emission tomography (FDG-PET) data of 137 patients with MCI and 80 healthy controls (HCs). The HC group data comes from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. The permutation test was used to compare the network parameters (characteristic path length, clustering coefficient, local efficiency, and global efficiency) between the two groups. Partial Pearson’s correlation analysis was used to calculate the correlations of the changes in gray matter volume and glucose intake in the key brain regions in MCI with the Alzheimer’s Disease Assessment Scale-Cognitive (ADAS-cog) sub-item scores.

Results: Significant changes in the brain network parameters (longer characteristic path length, larger clustering coefficient, and lower local efficiency and global efficiency) were greater in the structural network than in the metabolic network (longer characteristic path length) in MCI patients than in HCs. We obtained the key brain regions (left globus pallidus, right calcarine fissure and its surrounding cortex, left lingual gyrus) by scanning the hubs. The volume of gray matter atrophy in the left globus pallidus was significantly positively correlated with comprehension of spoken language (p = 0.024) and word-finding difficulty in spontaneous speech item scores (p = 0.007) in the ADAS-cog. Glucose intake in the three key brain regions was significantly negatively correlated with remembering test instructions items in ADAS-cog (p = 0.020, p = 0.014, and p = 0.008, respectively).

Conclusion: Structural brain networks showed more changes than metabolic brain networks in patients with MCI. Some brain regions with significant changes in betweenness centrality in both structural and metabolic networks were associated with MCI.

Multiple Connection Pattern Combination From Single-Mode Data for Mild Cognitive Impairment Identification

Mild cognitive impairment (MCI) is generally considered to be a key indicator for predicting the early progression of Alzheimer’s disease (AD). Currently, the brain connection (BC) estimated by fMRI data has been validated to be an effective diagnostic biomarker for MCI. Existing studies mainly focused on the single connection pattern for the neuro-disease diagnosis. Thus, such approaches are commonly insufficient to reveal the underlying changes between groups of MCI patients and normal controls (NCs), thereby limiting their performance. In this context, the information associated with multiple patterns (e.g., functional connectivity or effective connectivity) from single-mode data are considered for the MCI diagnosis. In this paper, we provide a novel multiple connection pattern combination (MCPC) approach to combine different patterns based on the kernel combination trick to identify MCI from NCs. In particular, sixty-three MCI cases and sixty-four NC cases from the ADNI dataset are conducted for the validation of the proposed MCPC method. The proposed method achieves 87.40% classification accuracy and significantly outperforms methods that use a single pattern.

An ALE Meta-Analysis of Specific Functional MRI Studies on Subcortical Vascular Cognitive Impairment

Background: Subcortical vascular cognitive impairment (sVCI), caused by cerebral small vessel disease, accounts for the majority of vascular cognitive impairment, and is characterized by an insidious onset and impaired memory and executive function. If not recognized early, it inevitably develops into vascular dementia. Several quantitative studies have reported the consistent results of brain regions in sVCI patients that can be used to predict dementia conversion. The purpose of the study was to explore the exact abnormalities within the brain in sVCI patients by combining the coordinates reported in previous studies.

Methods: The PubMed, Embase, and Web of Science databases were thoroughly searched to obtain neuroimaging articles on the amplitude of low-frequency fluctuation, regional homogeneity, and functional connectivity in sVCI patients. According to the activation likelihood estimation (ALE) algorithm, a meta-analysis based on coordinate and functional connectivity modeling was conducted.

Results: The quantitative meta-analysis included 20 functional imaging studies on sVCI patients. Alterations in specific brain regions were mainly concentrated in the frontal lobes including the middle frontal gyrus, superior frontal gyrus, medial frontal gyrus, and precentral gyrus; parietal lobes including the precuneus, angular gyrus, postcentral gyrus, and inferior parietal lobule; occipital lobes including the lingual gyrus and cuneus; temporal lobes including the fusiform gyrus and middle temporal gyrus; and the limbic system including the cingulate gyrus. These specific brain regions belonged to important networks known as the default mode network, the executive control network, and the visual network.

Conclusion: The present study determined specific abnormal brain regions in sVCI patients, and these brain regions with specific changes were found to belong to important brain functional networks. The findings objectively present the exact abnormalities within the brain, which help further understand the pathogenesis of sVCI and identify them as potential imaging biomarkers. The results may also provide a basis for new approaches to treatment.

Repeated anodal high-definition transcranial direct current stimulation over the left dorsolateral prefrontal cortex in mild cognitive impairment patients increased regional homogeneity in multiple brain regions

Transcranial direct current stimulation (tDCS) can improve cognitive function. However, it is not clear how high-definition tDCS (HD-tDCS) regulates the cognitive function and its neural mechanism, especially in individuals with mild cognitive impairment (MCI). This study aimed to examine whether HD-tDCS can modulate cognitive function in individuals with MCI and to determine whether the potential variety is related to spontaneous brain activity changes recorded by resting-state functional magnetic resonance imaging (rs-fMRI). Forty-three individuals with MCI were randomly assigned to receive either 10 HD-tDCS sessions or 10 sham sessions to the left dorsolateral prefrontal cortex (L-DLPFC). The fractional amplitude of low-frequency fluctuation (fALFF) and the regional homogeneity (ReHo) was computed using rs-fMRI data from all participants. The results showed that the fALFF and ReHo values changed in multiple areas following HD-tDCS. Brain regions with significant decreases in fALFF values include the Insula R, Precuneus R, Thalamus L, and Parietal Sup R, while the Temporal Inf R, Fusiform L, Occipital Sup L, Calcarine R, and Angular R showed significantly increased in their fALFF values. The brain regions with significant increases in ReHo values include the Temporal Inf R, Putamen L, Frontal Mid L, Precentral R, Frontal Sup Medial L, Frontal Sup R, and Precentral L. We found that HD-tDCS can alter the intensity and synchrony of brain activity, and our results indicate that fALFF and ReHo analysis are sensitive indicators for the detection of HD-tDCS during spontaneous brain activity. Interestingly, HD-tDCS increases the ReHo values of multiple brain regions, which may be related to the underlying mechanism of its clinical effects, these may also be related to a potential compensation mechanism involving the mobilization of more regions to complete a function following a functional decline.

Morphological, Structural, and Functional Networks Highlight the Role of the Cortical-Subcortical Circuit in Individuals With Subjective Cognitive Decline

Subjective cognitive decline (SCD) is considered the earliest stage of the clinical manifestations of the continuous progression of Alzheimer’s Disease (AD). Previous studies have suggested that multimodal brain networks play an important role in the early diagnosis and mechanisms underlying SCD. However, most of the previous studies focused on a single modality, and lacked correlation analysis between different modal biomarkers and brain regions. In order to further explore the specific characteristic of the multimodal brain networks in the stage of SCD, 22 individuals with SCD and 20 matched healthy controls (HCs) were recruited in the present study. We constructed the individual morphological, structural and functional brain networks based on 3D-T1 structural magnetic resonance imaging (sMRI), diffusion tensor imaging (DTI) and resting-state functional magnetic resonance imaging (rs-fMRI), respectively. A t-test was used to select the connections with significant difference, and a multi-kernel support vector machine (MK-SVM) was applied to combine the selected multimodal connections to distinguish SCD from HCs. Moreover, we further identified the consensus connections of brain networks as the most discriminative features to explore the pathological mechanisms and potential biomarkers associated with SCD. Our results shown that the combination of three modal connections using MK-SVM achieved the best classification performance, with an accuracy of 92.68%, sensitivity of 95.00%, and specificity of 90.48%. Furthermore, the consensus connections and hub nodes based on the morphological, structural, and functional networks identified in our study exhibited abnormal cortical-subcortical connections in individuals with SCD. In addition, the functional networks presented more discriminative connections and hubs in the cortical-subcortical regions, and were found to perform better in distinguishing SCD from HCs. Therefore, our findings highlight the role of the cortical-subcortical circuit in individuals with SCD from the perspective of a multimodal brain network, providing potential biomarkers for the diagnosis and prediction of the preclinical stage of AD.

Constructing Dynamic Brain Functional Networks via Hyper-Graph Manifold Regularization for Mild Cognitive Impairment Classification

Brain functional networks (BFNs) constructed via manifold regularization (MR) have emerged as a powerful tool in finding new biomarkers for brain disease diagnosis. However, they only describe the pair-wise relationship between two brain regions, and cannot describe the functional interaction between multiple brain regions, or the high-order relationship, well. To solve this issue, we propose a method to construct dynamic BFNs (DBFNs) via hyper-graph MR (HMR) and employ it to classify mild cognitive impairment (MCI) subjects. First, we construct DBFNs via Pearson's correlation (PC) method and remodel the PC method as an optimization model. Then, we use k-nearest neighbor (KNN) algorithm to construct the hyper-graph and obtain the hyper-graph manifold regularizer based on the hyper-graph. We introduce the hyper-graph manifold regularizer and the L1-norm regularizer into the PC-based optimization model to optimize DBFNs and obtain the final sparse DBFNs (SDBFNs). Finally, we conduct classification experiments to classify MCI subjects from normal subjects to verify the effectiveness of our method. Experimental results show that the proposed method achieves better classification performance compared with other state-of-the-art methods, and the classification accuracy (ACC), the sensitivity (SEN), the specificity (SPE), and the area under the curve (AUC) reach 82.4946 ± 0.2827%, 77.2473 ± 0.5747%, 87.7419 ± 0.2286%, and 0.9021 ± 0.0007, respectively. This method expands the MR method and DBFNs with more biological significance. It can effectively improve the classification performance of DBFNs for MCI, and has certain reference value for the research and auxiliary diagnosis of Alzheimer's disease (AD).

Characteristics of Multimodal Brain Connectomics in Patients With Schizophrenia and the Unaffected First-Degree Relatives

OBJECTIVE

Increasing pieces of evidence suggest that abnormal brain connectivity plays an important role in the pathophysiology of schizophrenia. As an essential strategy in psychiatric neuroscience, the research of brain connectivity-based neuroimaging biomarkers has gained increasing attention. Most of previous studies focused on a single modality of the brain connectomics. Multimodal evidence will not only depict the full profile of the brain abnormalities of patients but also contribute to our understanding of the neurobiological mechanisms of this disease.

METHODS

In the current study, 99 schizophrenia patients, 69 sex- and education-matched healthy controls, and 42 unaffected first-degree relatives of patients were recruited and scanned. The brain was parcellated into 246 regions and multimodal network analyses were used to construct brain connectivity networks for each participant.

RESULTS

Using the brain connectomics from three modalities as the features, the multi-kernel support vector machine method yielded high discrimination accuracies for schizophrenia patients (94.86%) and for the first-degree relatives (95.33%) from healthy controls. Using an independent sample (49 patients and 122 healthy controls), we tested the model and achieved a classification accuracy of 64.57%. The convergent pattern within the basal ganglia and thalamus-cortex circuit exhibited high discriminative power during classification. Furthermore, substantial overlaps of the brain connectivity abnormality between patients and the unaffected first-degree relatives were observed compared to healthy controls.

CONCLUSION

The current findings demonstrate that decreased functional communications between the basal ganglia, thalamus, and the prefrontal cortex could serve as biomarkers and endophenotypes for schizophrenia.

Constructing Dynamic Functional Networks via Weighted Regularization and Tensor Low-Rank Approximation for Early Mild Cognitive Impairment Classification

Brain functional networks constructed via regularization has been widely used in early mild cognitive impairment (eMCI) classification. However, few methods can properly reflect the similarities and differences of functional connections among different people. Most methods ignore some topological attributes, such as connection strength, which may delete strong functional connections in brain functional networks. To overcome these limitations, we propose a novel method to construct dynamic functional networks (DFN) based on weighted regularization (WR) and tensor low-rank approximation (TLA), and apply it to identify eMCI subjects from normal subjects. First, we introduce the WR term into the DFN construction and obtain WR-based DFNs (WRDFN). Then, we combine the WRDFNs of all subjects into a third-order tensor for TLA processing, and obtain the DFN based on WR and TLA (WRTDFN) of each subject in the tensor. We calculate the weighted-graph local clustering coefficient of each region in each WRTDFN as the effective feature, and use the t-test for feature selection. Finally, we train a linear support vector machine (SVM) classifier to classify the WRTDFNs of all subjects. Experimental results demonstrate that the proposed method can obtain DFNs with the scale-free property, and that the classification accuracy (ACC), the sensitivity (SEN), the specificity (SPE), and the area under curve (AUC) reach 87.0662% ± 0.3202%, 83.4363% ± 0.5076%, 90.6961% ± 0.3250% and 0.9431 ± 0.0023, respectively. We also achieve the best classification results compared with other comparable methods. This work can effectively improve the classification performance of DFNs constructed by existing methods for eMCI and has certain reference value for the early diagnosis of Alzheimer’s disease (AD).

MNL-Network: A Multi-Scale Non-local Network for Epilepsy Detection From EEG Signals

Epilepsy is a prevalent neurological disorder that threatens human health in the world. The most commonly used method to detect epilepsy is using the electroencephalogram (EEG). However, epilepsy detection from the EEG is time-consuming and error-prone work because of the varying levels of experience we find in physicians. To tackle this challenge, in this paper, we propose a multi-scale non-local (MNL) network to achieve automatic EEG signal detection. Our MNL-Network is based on 1D convolution neural network involving two specific layers to improve the classification performance. One layer is named the signal pooling layer which incorporates three different sizes of 1D max-pooling layers to learn the multi-scale features from the EEG signal. The other one is called a multi-scale non-local layer, which calculates the correlation of different multi-scale extracted features and outputs the correlative encoded features to further enhance the classification performance. To evaluate the effectiveness of our model, we conduct experiments on the Bonn dataset. The experimental results demonstrate that our MNL-Network could achieve competitive results in the EEG classification task.

Functional MRI-Specific Alterations in Executive Control Network in Mild Cognitive Impairment: An ALE Meta-Analysis

Background: Mild cognitive impairment (MCI) is regarded as a transitional stage between normal aging and Alzheimer's disease (AD) dementia. MCI individuals with deficits in executive function are at higher risk for progressing to AD dementia. Currently, there is no consistent result for alterations in the executive control network (ECN) in MCI, which makes early prediction of AD conversion difficult. The aim of the study was to find functional MRI-specific alterations in ECN in MCI patients by expounding on the convergence of brain regions with functional abnormalities in ECN.

Methods: We searched PubMed, Embase, and Web of Science to identify neuroimaging studies using methods including the amplitude of low frequency fluctuation/fractional amplitude of low-frequency fluctuation, regional homogeneity, and functional connectivity in MCI patients. Based on the Activation Likelihood Estimation algorithm, the coordinate-based meta-analysis and functional meta-analytic connectivity modeling were conducted.

Results: A total of 25 functional imaging studies with MCI patients were included in a quantitative meta-analysis. By summarizing the included articles, we obtained specific brain region changes, mainly including precuneus, cuneus, lingual gyrus, middle frontal gyrus, posterior cingulate cortex, and cerebellum posterior lobe, in the ECN based on these three methods. The specific abnormal brain regions indicated that there were interactions between the ECN and other networks.

Conclusions: This study confirms functional imaging specific abnormal markers in ECN and its interaction with other networks in MCI. It provides novel targets and pathways for individualized and precise interventions to delay the progression of MCI to AD.

Investigation on the Alteration of Brain Functional Network and Its Role in the Identification of Mild Cognitive Impairment

Mild cognitive impairment (MCI) is generally regarded as a prodromal stage of Alzheimer’s disease (AD). In coping with the challenges caused by AD, we analyzed resting-state functional magnetic resonance imaging data of 82 MCI subjects and 93 normal controls (NCs). The alteration of brain functional network in MCI was investigated on three scales, including global metrics, nodal characteristics, and modular properties. The results supported the existence of small worldness, hubs, and community structure in the brain functional networks of both groups. Compared with NCs, the network altered in MCI over all the three scales. In scale I, we found significantly decreased characteristic path length and increased global efficiency in MCI. Moreover, altered global network metrics were associated with cognitive level evaluated by neuropsychological assessments. In scale II, the nodal betweenness centrality of some global hubs, such as the right Crus II of cerebellar hemisphere (CERCRU2.R) and fusiform gyrus (FFG.R), changed significantly and associated with the severity and cognitive impairment in MCI. In scale III, although anatomically adjacent regions tended to be clustered into the same module regardless of group, discrepancies existed in the composition of modules in both groups, with a prominent separation of the cerebellum and a less localized organization of community structure in MCI compared with NC. Taking advantages of random forest approach, we achieved an accuracy of 91.4% to discriminate MCI patients from NCs by integrating cognitive assessments and network analysis. The importance of the used features fed into the classifier further validated the nodal characteristics of CERCRU2.R and FFG.R could be potential biomarkers in the identification of MCI. In conclusion, the present study demonstrated that the brain functional connectome data altered at the stage of MCI and could assist the automatic diagnosis of MCI patients.

Effective and Accurate Diagnosis of Subjective Cognitive Decline Based on Functional Connection and Graph Theory View

Subjective cognitive decline (SCD) is considered the earliest preclinical stage of Alzheimer’s disease (AD) that precedes mild cognitive impairment (MCI). Effective and accurate diagnosis of SCD is crucial for early detection of and timely intervention in AD. In this study, brain functional connectome (i.e., functional connections and graph theory metrics) based on the resting-state functional magnetic resonance imaging (rs-fMRI) provided multiple information about brain networks and has been used to distinguish individuals with SCD from normal controls (NCs). The consensus connections and the discriminative nodal graph metrics selected by group least absolute shrinkage and selection operator (LASSO) mainly distributed in the prefrontal and frontal cortices and the subcortical regions corresponded to default mode network (DMN) and frontoparietal task control network. Nodal efficiency and nodal shortest path showed the most significant discriminative ability among the selected nodal graph metrics. Furthermore, the comparison results of topological attributes suggested that the brain network integration function was weakened and network segregation function was enhanced in SCD patients. Moreover, the combination of brain connectome information based on multiple kernel-support vector machine (MK-SVM) achieved the best classification performance with 83.33% accuracy, 90.00% sensitivity, and an area under the curve (AUC) of 0.927. The findings of this study provided a new perspective to combine machine learning methods with exploration of brain pathophysiological mechanisms in SCD and offered potential neuroimaging biomarkers for diagnosis of early-stage AD.

Functional connectivity network estimation with an inter-similarity prior for mild cognitive impairment classification

Functional connectivity network (FCN) analysis is an effective technique for modeling human brain patterns and diagnosing neurological disorders such as Alzheimer's disease (AD) and its early stage, Mild Cognitive Impairment. However, accurately estimating biologically meaningful and discriminative FCNs remains challenging due to the poor quality of functional magnetic resonance imaging (fMRI) data and our limited understanding of the human brain. Inspired by the inter-similarity nature of FCNs, similar regions of interest tend to share similar connection patterns. Here, we propose a functional brain network modeling scheme by encoding Inter-similarity prior into a graph-regularization term, which can be easily solved with an efficient optimization algorithm. To illustrate its effectiveness, we conducted experiments to distinguish Mild Cognitive Impairment from normal controls based on their respective FCNs. Our method outperformed the baseline and state-of-the-art methods by achieving an 88.19% classification accuracy. Furthermore, post hoc inspection of the informative features showed that our method yielded more biologically meaningful functional brain connectivity.