Dataset Information

Hierarchical multivariate covariance analysis of metabolic connectivity.

ABSTRACT: Conventional brain connectivity analysis is typically based on the assessment of interregional correlations. Given that correlation coefficients are derived from both covariance and variance, group differences in covariance may be obscured by differences in the variance terms. To facilitate a comprehensive assessment of connectivity, we propose a unified statistical framework that interrogates the individual terms of the correlation coefficient. We have evaluated the utility of this method for metabolic connectivity analysis using [18F]2-fluoro-2-deoxyglucose (FDG) positron emission tomography (PET) data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) study. As an illustrative example of the utility of this approach, we examined metabolic connectivity in angular gyrus and precuneus seed regions of mild cognitive impairment (MCI) subjects with low and high β-amyloid burdens. This new multivariate method allowed us to identify alterations in the metabolic connectome, which would not have been detected using classic seed-based correlation analysis. Ultimately, this novel approach should be extensible to brain network analysis and broadly applicable to other imaging modalities, such as functional magnetic resonance imaging (MRI).

SUBMITTER: Carbonell F

PROVIDER: S-EPMC4269748 | biostudies-literature | 2014 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Hierarchical multivariate covariance analysis of metabolic connectivity.

Carbonell Felix F Charil Arnaud A Zijdenbos Alex P AP Evans Alan C AC Bedell Barry J BJ

Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 20141008 12

Conventional brain connectivity analysis is typically based on the assessment of interregional correlations. Given that correlation coefficients are derived from both covariance and variance, group differences in covariance may be obscured by differences in the variance terms. To facilitate a comprehensive assessment of connectivity, we propose a unified statistical framework that interrogates the individual terms of the correlation coefficient. We have evaluated the utility of this method for m ...[more]

PMID: 25294129

Similar Datasets

Project description:BACKGROUND: Identification of differentially expressed genes is a typical objective when analyzing gene expression data. Recently, Bayesian hierarchical models have become increasingly popular to solve this type of problems. These models show good performance in accommodating noise, variability and low replication of microarray data. However, the correlation between different fluorescent signals measured from a gene spot is ignored, which can diversely affect the data analysis step. In fact, the intensities of the two signals are significantly correlated across samples. The larger the log-transformed intensities are, the smaller the correlation is. RESULTS: Motivated by the complicated error relations in microarray data, we propose a multivariate hierarchical Bayesian framework for data analysis in the replicated microarray experiments. Gene expression data are modelled by a multivariate normal distribution, parameterized by the corresponding mean vectors and covariance matrixes with a conjugate prior distribution. Within the Bayesian framework, a generalized likelihood ratio test (GLRT) is also developed to infer the gene expression patterns. Simulation studies show that the proposed approach presents better operating characteristics and lower false discovery rate (FDR) than existing methods, especially when the correlation coefficient is large. The approach is illustrated with two examples of microarray analysis. The proposed method successfully detects significant genes closely related to the experimental states, which are verified by the biological information. CONCLUSIONS: The multivariate Bayesian model, compatible with the dependence between mean and variance in the univariate Bayesian model, relaxes the constant coefficient of variation assumption between measurements by adding a covariance structure. This model improves the identification of differentially expressed genes significantly since the Bayesian model fit well with the microarray data.

Project description:BackgroundPatients with bipolar disorder (BD) and major depressive disorder (MDD) exhibit depressive episodes with similar symptoms despite having different and poorly understood underlying neurobiology, often leading to misdiagnosis and improper treatment. This exploratory study examined whole-brain functional connectivity (FC) using FC multivariate pattern analysis (fc-MVPA) to identify the FC patterns with the greatest ability to distinguish between currently depressed patients with BD type I (BD I) and those with MDD.MethodologyIn a cross-sectional design, 41 BD I, 40 MDD patients and 63 control participants completed resting state functional magnetic resonance imaging scans. Data-driven fc-MVPA, as implemented in the CONN toolbox, was used to identify clusters with differential FC patterns between BD patients and MDD patients. The identified cluster was used as a seed in a post hoc seed-based analysis (SBA) to reveal associated connectivity patterns, followed by a secondary ROI-to-ROI analysis to characterize differences in connectivity between these patterns among BD I patients, MDD patients and controls.ResultsFC-MVPA identified one cluster located in the right frontal pole (RFP). The subsequent SBA revealed greater FC between the RFP and posterior cingulate cortex (PCC) and between the RFP and the left inferior/middle temporal gyrus (LI/MTG) and lower FC between the RFP and the left precentral gyrus (LPCG), left lingual gyrus/occipital cortex (LLG/OCC) and right occipital cortex (ROCC) in MDD patients than in BD patients. Compared with the controls, ROI-to-ROI analysis revealed lower FC between the RFP and the PCC and greater FC between the RFP and the LPCG, LLG/OCC and ROCC in BD patients; in MDD patients, the analysis revealed lower FC between the RFP and the LLG/OCC and ROCC and greater FC between the RFP and the LI/MTG.ConclusionsDifferences in the RFP FC patterns between currently depressed patients with BD and those with MDD suggest potential neuroimaging markers that should be further examined. Specifically, BD patients exhibit increased FC between the RFP and the motor and visual networks, which is associated with psychomotor symptoms and heightened compensatory frontoparietal FC to counter distractibility. In contrast, MDD patients exhibit increased FC between the RFP and the default mode network, corresponding to sustained self-focus and rumination.

Dataset Information

Hierarchical multivariate covariance analysis of metabolic connectivity.

Publications

Hierarchical multivariate covariance analysis of metabolic connectivity.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets