Abstract not found

Abstract not found

Unveiling the modular structure of biological networks can reveal important organizational patterns in the cell. Many graph partitioning algorithms have been proposed towards this end. However, most approaches only consider a single, optimal decomposition of the network. In this work, we make use of the multitude of near-optimal clusterings in order to explore the dynamics of network clusterings and how those dynamics relate to the structure of the underlying network. We recast the modularity optimization problem as an integer linear program with diversity constraints. These constraints produce an ensemble of dissimilar but still highly modular clusterings. We apply our approach to four social and biological networks and show how optimal and near-optimal solutions can be used in conjunction to identify deeper community structure in the network, including inter-community dynamics, communities that are especially resilient to change, and core-and-peripheral community members. 1.

An active area of neuroimaging research involves examining functional relationships between spatially remote brain regions. When determining whether two brain regions exhibit significant correlation due to true functional connectivity, one must account for the background spatial correlation inherent in neuroimaging data. We define background correlation as spatiotemporal correlation in the data caused by factors other than neurophysiologically based functional associations such as scanner induced correlations and image preprocessing. We develop a 4D spatiotemporal wavelet packet resampling method which generates surrogate data that preserves only the average background spatial correlation within an axial slice, across axial slices, and through each voxel time series, while excluding the specific correlations due to true functional relationships. We also extend an amplitude adjustment algorithm which adjusts our surrogate data to closely match the amplitude distribution of the original data. Our method improves upon existing wavelet-based methods and extends them to 4D. We apply our resampling technique to determine significant functional connectivity from resting state and motor task fMRI datasets.

Functional neuroimaging studies provide converging evidence for existence of intrinsic brain networks activated during resting states and deactivated with selective cognitive demands. Whether task-related deactivation of the default mode network signifies depressed activity relative to the remaining brain or simply lower activity relative to its resting state remains controversial. We employed 3D arterial spin labeling imaging to examine regional cerebral blood flow (CBF) during rest, a spatial working memory task, and a second rest. Change in regional CBF from rest to task showed significant normalized and absolute CBF reductions in posterior cingulate, posterior-inferior precuneus, and medial frontal lobes. A Statistical Parametric Mapping connectivity analysis, with an a priori seed in the posterior cingulate cortex, produced deactivation connectivity patterns consistent with the classic ‘‘default mode network’ ’ and activation connectivity anatomically consistent

We propose maximum covariance unfolding (MCU), a manifold learning algorithm for simultaneous dimensionality reduction of data from different input modalities. Given high dimensional inputs from two different but naturally aligned sources, MCU computes a common low dimensional embedding that maximizes the cross-modal (inter-source) correlations while preserving the local (intra-source) distances. In this paper, we explore two applications of MCU. First we use MCU to analyze EEG-fMRI data, where an important goal is to visualize the fMRI voxels that are most strongly correlated with changes in EEG traces. To perform this visualization, we augment MCU with an additional step for metric learning in the high dimensional voxel space. Second, we use MCU to perform cross-modal retrieval of matched image and text samples from Wikipedia. To manage large applications of MCU, we develop a fast implementation based on ideas from spectral graph theory. These ideas transform the original problem for MCU, one of semidefinite programming, into a simpler problem in semidefinite quadratic linear programming. 1

Abstract—In medicine, the nature of an illness is often determined through behavioral or biological markers. The process of diagnosis becomes difficult when dealing with mental disorders since they rely primarily on behavioral markers. Schizophrenia is an example of a complex mental disorder that relies on aberrant behavior such as auditory hallucinations, dampening of emotions, paranoia, etc. This research is an attempt to determine a biological marker for schizophrenia through the use of functional magnetic resonance imaging (fMRI). In this paper, we propose a method of classification of schizophrenia and healthy controls, using a neural network approach and functional brain ‘modes’ estimated from resting state data using independent component analysis. A reliable technique for discriminating schizophrenia based upon fMRI would be a significant advance and may also provide additional information about the biological implications of mental illness. A

Contents lists available at ScienceDirect

www.elsevier.com/locate/ynimg

doi: 10.3389/fnsys.2010.00020 The relation of ongoing brain activity, evoked neural responses, and cognition