A novel node-level structure embedding and alignment representation of structural networks for brain disease analysis
Graphical abstract
Introduction
Network analysis techniques can characterize brain structures and activities from the perspective of the complex system (Meyer-Lindenberg, 2010). Especially, with developments in noninvasive neuroimaging technologies, such as diffusion tensor image (DTI) and functional magnetic resonance imaging (fMRI), the brain networks based on the various neuroimaging data have been applied to the analysis of brain structures and functions. Existing studies have suggested that the structure of the brain network is related to cognitive states (Sharp, Scott, Leech, 2014, Wang et al., 2020), the nervous system diseases (Beaty, Benedek, Silvia, Schacter, 2016, Huang et al., 2020), and the injury of brain (Bullmore and Sporns, 2012).
In the brain network, nodes denote the brain regions, and edges correspond to the physical connections or pairwise similarity of rs-fMRI time series. To date, most researchers usually use some topological measures to characterize brain networks, and these studies mainly focus on three levels: (1) edge level, where the brain networks are described by using connections between nodes, such as connectivity strengths, subgraphs, and motifs (Zhang et al., 2018). (2) node level, where the brain networks are charactered by using the local topological structure of nodes, such as node degrees, local clustering coefficients, and betweenness (Rubinov and Sporns, 2010). (3) network level, where the brain networks are represented by using whole properties of the network, such as small-world characteristics. However, in the edge-level, it is difficult to design an efficient model for identifying brain diseases since a large number of variables (i.e., connections) (Yu et al., 2017). In addition, connections of brain network are usually considered as independent variables in most edge-level based methods, ignoring the relationships among them. Losing such relationships may decrease the performance of brain disease identification. On the other hand, in the network-level based methods, the local network properties are usually neglected, thus leading to less sensitive in capturing the details of the structural changes caused by brain diseases.
Compared to the edge-level based and network-level based methods, node-level based methods provide a proper scale of brain network representations (Gad Elkarim et al., 2012). Generally, node-level based methods characterize the structure of brain networks by integrating connectivity information (Wang et al., 2010). For example, the node degree is calculated by the number of edges linked to the node. It reflects the information of the local neighboring node, thus characterizing the local topological details of networks. Besides, using node-level based methods can directly identify disease-related brain regions, which is critical for understanding brain diseases (Shen et al., 2013).
The node-level measurements have been widely applied to group-level brain network analysis. For example, Hadley et al. (2016) utilized measurements of functional segregation and integration to analysis 32 unmedicated patients with schizophrenia and 32 matched healthy controls, and found reduced clustering and increased global efficiency in the schizophrenia. Liao et al. (2010) observed smaller degree of connectivity in patients of mesial temporal lobe epilepsy. Moreover, node-level measurements have been applied to identify brain diseases. For example, the local clustering coefficient has been used for mild cognitive impairment classification in several studies (Jie, Zhang, Gao, Wang, Wee, Shen, 2014, Wee, Yap, Zhang, Denny, Browndyke, Potter, Welsh-Bohmer, Wang, Shen, 2012). However, there are two disadvantages in existing studies. Firstly, these node-level measurements are often calculated based on the graph theory that characterizes node structure information with a single type of properties, and thus the measurement of each node is usually represented as a single value. As a result, the important information, such as small changes of the network, cannot be accurately characterized. These two brain networks in Fig. 1 have small structural differences. Obviously, the traditional node-level measurement (i.e., node degree) can not reflect the real structural differences of these networks. Secondly, most node measurements are usually calculated based on the local structure of nodes, ignoring the larger scale of network structure. However, existing studies found that larger scales of network structure can increase the interpretability of the hierarchy of the brain, which may help identify disease-related biomarkers and further improve the diagnosis performance of brain diseases (Tomson, Schreiner, Narayan, Rosser, Enrique, Silva, Allen, Bookheimer, Bearden, 2015, Wang, Huang, Liu, Zhang, 2019). Therefore, designing an appropriate node-level representation to characterize structure information of brain networks is still a critical and challenging task.
Recently, network embedding methods have been applied to information networks, such as social networks (Zhang et al., 2017) and language networks (Chen and Sun, 2017). These embedding methods can be categorized into two broad categories: Factorization based methods and Random Walk based methods. Briefly, the former category deals with a matrix about the connection information (e.g. the connection matrix or the Laplace matrix), and factorize this matrix to obtain the vector-based representation of nodes (Tang et al., 2015). The latter category deals with node sequences generated by using fixed-length random walks in the network. The vector-based representations are obtained by maximizing the probability of observing the last node and the next node (Perozzi et al., 2014). These methods can preserve higher-order proximity between nodes, and achieve meaningful results in the task of link predictions and node classifications. However, these methods are typically designed to the single network. Hence, it cannot be directly used in the group-level analysis due to that node vectors representations from different networks are not in the common space (Faruqui and Dyer, 2014). A feasible method to solve this problem is to learn the representation of whole networks, but classification performance of this method is not ideal since local structural changes from disease-related regions are masked when learning the global network representation (Haroon, Miller, Sanacora, 2017, Huang et al., 2019).
To overcome these problems, in this paper, we define a novel node-level structure embedding and alignment (nSEA) representation of brain networks. The main idea of our method is to map node-level structure information into a vector representation with the common space. Specifically, we first perform the skip-gram model (Levy and Goldberg, 2014) to obtain node vector representations that contain richer structure information of nodes. Then, we align the vector representations of all brain networks into the common space by optimizing the orthogonal Procrustes problem (Heusser et al., 2017). Furthermore, we develop an nSEA representation based learning (nSEAL) framework for two group-level brain network analyses, including a statistic analysis and classifications. The statistic analysis is performed on node distance vectors generated by the vector representation of each node. In the classification, we compute the pairwise distance between brain networks by jointing node vector representations and node degree vectors, and then use the k nearest neighbor (KNN) classifier to identify the label of each network. Experiment results on a real schizophrenia dataset demonstrate that our proposed method can not only outperform several state-of-the-art approaches in the brain disease diagnosis, but also discover disease-related regions for improving the understanding of schizophrenia pathology.
The contributions of our proposed method can be summarized into the following three folds: (1) we define an nSEA representation of brain networks to convey the richer node-level structural information than traditional node measurements. (2) We develop an nSEAL framework for brain disease classifications and a statistic analysis. (3) We evaluate the performances of the proposed nSEAL method on a real schizophrenia dataset.
The rest of the paper is organized as follows. In Section 2, we describe the materials used in this study and present our proposed method and learning framework. Then, we introduce experiments and results in Section 3. In Section 4, we give the discussions for the experiments and results. Finally, we conclude this paper in Section 5.
Section snippets
Materials and methods
Fig. 2 illustrates our proposed nSEAL framework for brain disease analysis, which consists of four main components, i.e., image pre-processing and brain network construction, nSEA representation learning, node distance computation for the statistic analysis, network distance computation for the brain disease classification. Next, we will give the details of each component in our proposed nSEAL framework.
Experimental setting
To evaluate the effectiveness of our proposed method, we perform a series of experiments on the real brain network data. Specifically, we first perform the node distance analysis, and then measure the group difference based on the network distance. Further, two types of brain networks (i.e., structural brain network and functional brain network) datasets are used in classification experiments.
Node distance analysis
In this subsection, we first present node distances between a individual network and centroid templates
Discussion
In this paper, we propose a novel node-level structural embedding and alignment representation of brain networks. Different from conventional brain network representation methods that measure the specific type of node structural information with a single value, our proposed nSEA representation method can characterize the topological properties of nodes as feature vectors, thus can accurately characterize topological information of brain regions. The experimental results on a real schizophrenia
Conclusion
In this paper, we propose a novel node-level structure embedding and alignment representation of brain networks. Different from the conventional brain network representation methods that usually use a single value to measure the specific type of structural information, our proposed nSEA representation method can capture the topological properties of nodes as feature vectors, thus can more accurately characterize structural information of brain networks. Furthermore, we develop an nSEA
CRediT authorship contribution statement
Jiashuang Huang: Conceptualization, Methodology, Writing - original draft. Mingliang Wang: Writing - review & editing. Xijia Xu: Data curation. Biao Jie: Writing - review & editing. Daoqiang Zhang: Supervision, Writing - review & editing.
Declaration of Competing Interest
Authors declare that they have no conflict of interest.
Acknowledgments
This study was supported by the National Key Research and Development Program of China (Nos. 2018YFC2001600, 2018YFC2001602, 2018ZX10201002, 2016YFC1306900) and the National Natural Science Foundation of China (Nos. 61861130366, 61732006, 61976006, 61876082, 81771444), the Royal Society-Academy of Medical Sciences Newton Advanced Fellowship (No. NAFnR1n180371), the 16th Batch of Six Talent Peak Projects in Jiangsu (Nos. WSN-166), the Key Research and Development Plan of Jiangsu (Nos. BE2019707).
References (60)
- et al.
Multivariate brain network graph identification in functional MRI
Med. Image Anal.
(2017) - et al.
Creative cognition and brain network dynamics
Trends Cogn. Sci.
(2016) - et al.
Functional connectivity dynamically evolves on multiple time-scales over a static structural connectome: models and mechanisms
NeuroImage
(2017) - et al.
Loss and altered spatial distribution of oligodendrocytes in the superior frontal gyrus in schizophrenia
Biol. Psychiatry
(2003) - et al.
Gray matter volume reduction in rostral middle frontal gyrus in patients with chronic schizophrenia
Schizophr. Res.
(2010) - et al.
Deep neural network with weight sparsity control and pre-training extracts hierarchical features and enhances classification performance: evidence from whole-brain resting-state functional connectivity patterns of schizophrenia
Neuroimage
(2016) - et al.
Network community structure alterations in adult schizophrenia: identification and localization of alterations
NeuroImage
(2016) - et al.
Dysfunctional supplementary motor area implication during attention and time estimation tasks in schizophrenia: a PET-O15 water study
Neuroimage
(2005) - et al.
Complex network measures of brain connectivity: uses and interpretations
Neuroimage
(2010) - et al.
Groupwise whole-brain parcellation from resting-state fMRI data for network node identification
Neuroimage
(2013)
Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain
Neuroimage
Identification of MCI individuals using structural and functional connectivity networks
Neuroimage
Volumetric analysis of sulci/gyri-defined in vivo frontal lobe regions in schizophrenia: precentral gyrus, cingulate gyrus, and prefrontal region
Psychiatry Res.
Extracting semantic representations from word co-occurrence statistics: stop-lists, stemming, and SVD
Behav. Res. Methods
The economy of brain network organization
Nat. Rev. Neurosci.
t-BNE: tensor-based brain network embedding
Proceedings of the 2017 SIAM International Conference on Data Mining
Mining brain networks using multiple side views for neurological disorder identification
2015 IEEE International Conference on Data Mining
Training SVM with indefinite kernels
Proceedings of the 25th international conference on Machine learning
Task-guided and path-augmented heterogeneous network embedding for author identification
Proceedings of the Tenth ACM International Conference on Web Search and Data Mining
Convolutional neural networks on graphs with fast localized spectral filtering
Advances in Neural Information Processing Systems
The human brainnetome atlas: a new brain atlas based on connectional architecture
Cereb. Cortex
Improving vector space word representations using multilingual correlation
Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics
Medinria: DT-MRI processing and visualization software
Similar Tensor Workshop
A framework for quantifying node-level community structure group differences in brain connectivity networks
International Conference on Medical Image Computing and Computer-Assisted Intervention
Procrustes Problems
Key functional circuitry altered in schizophrenia involves parietal regions associated with sense of self
Hum. Brain Mapp.
Change in brain network topology as a function of treatment response in schizophrenia: a longitudinal resting-state fMRI study using graph theory
NPJ Schizophr.
Inflammation, glutamate, and glia: a trio of trouble in mood disorders
Neuropsychopharmacology
Hypertools: a python toolbox for gaining geometric insights into high-dimensional data
J. Mach. Learn. Res.
Cited by (12)
Modeling dynamic characteristics of brain functional connectivity networks using resting-state functional MRI
2021, Medical Image AnalysisCitation Excerpt :Recent advances in network science have allowed new insights in understanding the brain from a system perspective. Characterizing brain networks derived from resting-state functional magnetic resonance imaging (rs-fMRI) data has not only pointed out important connectome features of brain functioning, but also led to the development of biomarkers quantifying re-organizational mechanisms of brain diseases (Sporns, 2013; Stam, 2014; Huang et al., 2020). Generally, a brain network (a.k.a. , connectome) can be characterized by a set of nodes and edges, where nodes represent the predefined regions of interest (ROIs) in the brain and edges denote either physical connections (structure networks) or statistical dependence relationships between the time series of measured neurophysiological signals (functional networks) (Fornito et al., 2016).
Beta-informativeness-diffusion multilayer graph embedding for brain network analysis
2024, Frontiers in NeuroscienceMulti-Discriminator Active Adversarial Network for Multi-Center Brain Disease Diagnosis
2023, IEEE Transactions on Big DataSelf-Supervised Federated Adaptation for Multi-Site Brain Disease Diagnosis
2023, IEEE Transactions on Big Data