Abstract

The analysis of the brain from a connectivity perspective is revealing novel insights into brain structure and function. Discovery is, however, hindered by the lack of prior knowledge used to make hypotheses. Additionally, exploratory data analysis is made complex by the high dimensionality of data. Indeed, to assess the effect of pathological states on brain networks, neuroscientists are often required to evaluate experimental effects in case-control studies, with hundreds of thousands of connections. In this paper, we propose an approach to identify the multivariate relationships in brain connections that characterize two distinct groups, hence permitting the investigators to immediately discover the subnetworks that contain information about the differences between experimental groups. In particular, we are interested in data discovery related to connectomics, where the connections that characterize differences between two groups of subjects are found. Nevertheless, those connections do not necessarily maximize the accuracy in classification since this does not guarantee reliable interpretation of specific differences between groups. In practice, our method exploits recent machine learning techniques employing sparsity to deal with weighted networks describing the whole-brain macro connectivity. We evaluated our technique on functional and structural connectomes from human and murine brain data. In our experiments, we automatically identified disease-relevant connections in datasets with supervised and unsupervised anatomy-driven parcellation approaches and by using high-dimensional datasets.

Highlights

  • Experiments with connectomes are typically designed by comparing a studied group with a control group to identify brain-network topological biomarkers relevant to the studied group[4]

  • The main drawback of the aforementioned approaches is the limited interpretability of the graph statistics, as they miss the local characterization of the groups in terms of the differences in the connectivity; instead, they employ global statistics that are difficult to translate into clinical settings for local analysis

  • We are interested in data discovery related to connectomics, where the connections that characterize differences between two groups of subjects are found, and where maximizing accuracy does not guarantee reliable interpretation since similar accuracies can be obtained from distinct sets of features[25]

Read more

Summary

Introduction

Experiments with connectomes are typically designed by comparing a studied group with a control group to identify brain-network topological biomarkers relevant to the studied group[4]. A method that allows insights on local connectivity patterns in case-control studies relies on network-based statistics (NBS) In this approach, the connectivity between pairs of brain regions is tested for significance using univariate statistics for functional[15] and anatomical[16] connectivity disturbances. Our method directly performs a sparse version of linear discriminant analysis (LDA) that, by design, tries to optimize the feature selection step aimed at discriminating the groups This approach allows the proposed method to be more specific in terms of identified discriminating connections. The proposed model is based on an ensemble of sparse linear discriminant models allowing to find the networks’ elements (a set of edges) that can consistently distinguish two groups, in the attempt to minimize the subset of selected connectivity features and simultaneously maximize the difference between the groups[30]. Recent studies have shown that a sparsity-based approach can be useful in graph/connectome analysis, as they can highlight significant connections when prior knowledge is missing[24,31,32,33]

Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call