Brain network analysis has been widely applied in neuroimaging studies. A hyper-network construction method was previously proposed to characterize the high-order relationships among multiple brain regions, where every edge is connected to more than two brain regions and can be represented by a hyper-graph. A brain functional hyper-network is constructed by a sparse linear regression model using resting-state functional magnetic resonance imaging (fMRI) time series, which in previous studies has been solved by the lasso method. Despite its successful application in many studies, the lasso method has some limitations, including an inability to explain the grouping effect. That is, using the lasso method may cause relevant brain regions be missed in selecting related regions. Ideally, a hyper-edge construction method should be able to select interacting brain regions as accurately as possible. To solve this problem, we took into account the grouping effect among brain regions and proposed two methods to improve the construction of the hyper-network: the elastic net and the group lasso. The three methods were applied to the construction of functional hyper-networks in depressed patients and normal controls. The results showed structural differences among the hyper-networks constructed by the three methods. The hyper-network structure obtained by the lasso was similar to that obtained by the elastic net method but very different from that obtained by the group lasso. The classification results indicated that the elastic net method achieved better classification results than the lasso method with the two proposed methods of hyper-network construction. The elastic net method can effectively solve the grouping effect and achieve better classification performance.

Despite many studies reporting hemispheric asymmetry in the representation and processing of emotions, the essence of the asymmetry remains controversial. Brain network analysis based on electroencephalography (EEG) is a useful biological method to study brain function. Here, EEG data were recorded while participants watched different emotional videos. According to the videos’ emotional categories, the data were divided into four categories: high arousal high valence (HAHV), low arousal high valence (LAHV), low arousal low valence (LALV) and high arousal low valence (HALV). The phase lag index as a connectivity index was calculated in theta (4–7 Hz), alpha (8–13 Hz), beta (14–30 Hz) and gamma (31–45 Hz) bands. Hemispheric networks were constructed for each trial, and graph theory was applied to quantify the hemispheric networks’ topological properties. Statistical analyses showed significant topological differences in the gamma band. The left hemispheric network showed significantly higher clustering coefficient (C_p), global efficiency (E_g) and local efficiency (E_loc) and lower characteristic path length (L_p) under HAHV emotion. The right hemispheric network showed significantly higher C_p and E_loc and lower L_p under HALV emotion. The results showed that the left hemisphere was dominant for HAHV emotion, while the right hemisphere was dominant for HALV emotion. The research revealed the relationship between emotion and hemispheric asymmetry from the perspective of brain networks.

Recent works have shown that the resting-state brain functional connectivity hypernetwork, where can be connected multiple nodes, are effective technique for brain disease diagnosis and classification research. The lasso method was used to construct hypernetworks by solving sparse linear regression models in previous research. But, constructing a hypernetwork based on the lasso method simply selects a single variable, in that it lacks the ability to interpret the grouping effect. Considering the group structure problem, the previous study proposed to create a hypernetwork based on the elastic net and the group lasso methods, and the results showed that the former method had best classification performance. However, the highly correlated variables selected by the elastic net method were not necessarily in the active set in the group. Therefore, we extended our research to address this issue. Herein, we propose a new method that introduces the sparse group lasso method to improve the construction of the hypernetwork by solving the group structure problem of the brain regions. We used the traditional lasso, group lasso method, and sparse group lasso method to construct a hypernetwork in patients with depression and normal subjects. Meanwhile, another clustering coefficients (clustering coefficients based on pairs of nodes) was also introduced to extract feature with traditional clustering coefficients. Two types of features with significant differences obtained after feature selection were subjected to multi-kernel learning for feature fusion and classification using each method respectively. The network topology results revealed differences among the three networks, where hypernetwork using the lasso method was the strictest; the group lasso, most lenient; and the sgLasso method, moderate. But the network topology of the sparse group lasso method was similar to that of the group lasso method but different from the lasso method. The classification results show that the sparse group lasso method achieves the best classification accuracy by using multi-kernel learning technique, which indicates that better classification performance can be achieved when the group structure exists and the group structure is properly extended.

High-order functional connectivity networks are rich in time information that can reflect dynamic changes in functional connectivity between brain regions. Accordingly, such networks are widely used to classify brain diseases. However, traditional methods for processing high-order functional connectivity networks generally include the clustering method, which reduces data dimensionality. As a result, such networks cannot be effectively interpreted in the context of neurology. Additionally, due to the large scale of high-order functional connectivity networks, it can be computationally very expensive to use complex network or graph theory to calculate certain topological properties. Here, we propose a novel method of generating a high-order minimum spanning tree functional connectivity network. This method increases the neurological significance of the high-order functional connectivity network, reduces network computing consumption, and produces a network scale that is conducive to subsequent network analysis. To ensure the quality of the topological information in the network structure, we used frequent subgraph mining technology to capture the discriminative subnetworks as features and combined this with quantifiable local network features. Then we applied a multikernel learning technique to the corresponding selected features to obtain the final classification results. We evaluated our proposed method using a data set containing 38 patients with major depressive disorder and 28 healthy controls. The experimental results showed a classification accuracy of up to 97.54%.

In recent years, functional brain network topological features have been widely used as classification features. Previous studies have found that network node scale differences caused by different network parcellation definitions significantly affect the structure of the constructed network and its topological properties. However, we still do not know how network scale differences affect the classification accuracy, performance of classification features, and effectiveness of the feature selection strategy using P values in terms of the machine learning method. This study used five scale parcellations, involving 90, 256, 497, 1003, and 1501 nodes. Three local properties of resting-state functional brain networks were selected (degree, betweenness centrality, and nodal efficiency), and the support vector machine method was used to construct classifiers to identify patients with major depressive disorder. We analyzed the impact of the five scales on classification accuracy. In addition, the effectiveness and redundancy of features obtained by the different scale parcellations were compared. Finally, traditional statistical significance (P value) was verified as a feature selection criterion. The results showed that the feature effectiveness of different scales was similar; in other words, parcellation with more regions did not provide more effective discriminative features. Nevertheless, parcellation with more regions did provide a greater quantity of discriminative features, which led to an improvement in the accuracy of the classification. However, due to the close distance between brain regions, the redundancy of parcellation with more regions was also greater. The traditional P value feature selection strategy is feasible with different scales, but our analysis showed that the traditional P<0.05 threshold was too strict for feature selection. This study provides an important reference for the selection of network scales when applying topological properties of brain networks to machine learning methods.