Dynamic functional connectivity analysis is an effective way to capture the networks that are functionally associated and continuously changing over scanning period. However, these methods mostly analyze dynamic associations across activation patterns of spatial while assuming stationary. Hence, a model allows for variability in both domains reduces assumptions imposed on data provides extracting spatiotemporal networks. Independent vector (IVA) joint blind source separation technique...

Dynamic functional network connectivity (dFNC) analysis is a widely-used to study associations between dynamic correlations and cognitive abilities. Traditional methods analyze time-varying association of different spatial networks while assuming that the itself stationary. However, there has been very little work focused on voxelwise variability. Exploiting variability across both temporal domains provide more promising direction obtain reliable patterns. for extracting spatio-temporal...

The extraction of common and distinct biomedical signatures among different populations allows for a more detailed study the group-specific as well information populations. A number subspace analysis algorithms have been developed successfully applied to data fusion, however they are limited joint only couple datasets. Since is very promising multi-subject medical imaging well, we focus on this problem propose new method based independent vector (IVA) (IVA-CS) analysis. IVA-CS leverages...

Abstract Data‐driven methods have been widely used in functional magnetic resonance imaging (fMRI) data analysis. They extract latent factors, generally, through the use of a simple generative model. Independent component analysis (ICA) and dictionary learning (DL) are two popular data‐driven that based on different forms diversity—statistical properties data—statistical independence for ICA sparsity DL. Despite their popularity, comparative advantage emphasizing one property over another...

Methods based on independent component analysis (ICA) and canonical correlation (CCA) as well their various extensions have become popular for the fusion of multimodal data they minimize assumptions about relationships among multiple datasets. Two important that are widely used, joint ICA (jICA) parallel (pICA), make a number simplifying might limit usefulness such identical mixing matrices jICA, requirement same components jICA pICA. In this paper, we propose new, flexible hybrid method...

The existence of complementary information across multiple sensors has driven the proliferation multivariate datasets. Exploitation this common information, while minimizing assumptions imposed on data led to popularity data-driven methods. Independent vector analysis (IVA), in particular, provides a flexible and effective approach for fusion data. In many practical applications, important prior about exists incorporating into IVA model is expected yield improved separation performance....

Characterizing time-evolving networks is a challenging task, but it crucial for understanding the dynamic behavior of complex systems such as brain. For instance, how spatial functional connectivity in brain evolve during task not well-understood. A traditional approach neuroimaging data analysis to make simplifications through assumption static networks. In this paper, without assuming time and/or space, we arrange temporal higher-order tensor and use fac-torization model called PARAFAC2...

Due to the potential for object occlusion in crowded areas, use of multiple cameras video surveillance has prevailed over a single camera. This motivated development number techniques analyze such multi-camera sequences. However, most these require camera calibration step, which is cumbersome and must be done every new configuration. Additionally, fail exploit complementary information across datasets. We propose data-driven solution problem by making inherent similarity temporal signatures...

The automated detection of abandoned objects is a quickly developing and widely researched field in video processing with specific application to surveillance. In the recent years, number approaches have been proposed automatically detect objects. However, these techniques require prior knowledge certain properties object such as its shape color, classify foreground object. performance tracking-based degrades complex scenes, i.e., when occluded or case crowding. this paper, we propose...

The success of many joint blind source separation techniques is dependent upon accurate estimation the common signal subspace order across multiple datasets. This has stimulated development to estimate number signals two datasets, in particular, a method that uses information theoretic criteria using canonical correlation coefficients likelihood formulation and based stage procedure, principal component analysis analysis. However, these methods are limited In this paper, we propose on...

There is a growing need for flexible methods the analysis of large-scale functional magnetic resonance imaging (fMRI) data estimation global signatures that summarize population while preserving individual-specific traits. Independent vector (IVA) data-driven method jointly estimates spatio-temporal patterns from multi-subject fMRI data, and effectively preserves subject variability. However, as we show, IVA performance negatively affected when number datasets components increases especially...

Automated detection of abandoned object (AO) is an important application in video surveillance for security purposes. Because its importance, a number techniques have been proposed to automatically detect objects the past years. However, these require prior knowledge on properties such as shape and color, order classify foreground object. In contrast, independent component analysis (ICA) does not knowledge. it can only model one dataset at time, thus limiting usage monochrome frames. this...

Aim: In this work, we propose the novel use of adaptively constrained independent vector analysis (acIVA) to effectively capture temporal and spatial properties dynamic blood-oxygen-level-dependent (BOLD) activity (dBA), efficiently quantify property dBA (sdBA). We also incorporate into study brain dynamics gain insight activity-connectivity co-evolution patterns. Introduction: Studies human using functional magnetic resonance imaging (fMRI) have enabled identification unique network...

Dynamic functional connectivity (dFC) analysis enables us to capture the time-varying interactions between brain regions and can lead powerful biomarkers. Most dFC studies are focused on study of temporal dynamics require significant post-processing summarize results analysis. In this paper, we introduce an effective framework that makes use independent vector (IVA) with fractional amplitude low frequency fluctuation (fALFF) features extracted from task magnetic resonance imaging (fMRI)...

Data-driven methods based on independent component analysis (ICA) and its extensions, have been attractive for data fusion as they minimize the assumptions placed data. Two widely used extensions of ICA, joint ICA (jICA) multiset canonical correlation prior to (MCCA-jICA) fuse from different datasets by assuming identical mixing matrices. However, these typically only take common features into account within linked disregarding available distinct features, thus limiting their usefulness. In...

In application to functional magnetic resonance imaging (fMRI) data analysis, a number of fusion algorithms have shown success in extracting interpretable brain networks that can distinguish two groups such populations-patients with mental disorder and the healthy controls. However, there are situations where more than exist as multi-task fMRI data. Therefore, this work we propose use IVA effectively extract information is able across multiple when applied fusion. The performance...

The detection of steady state visual evoked potentials (SSVEPs) has been identified as an effective solution for brain computer interface (BCI) systems well neurocognitive investigations visually related tasks. SSVEPs are induced at the same frequency stimuli and can be observed in scalp-based recordings electroencephalogram signals, though they one component buried amongst normal signals complex noise. Variations individual response latencies presence multiple biological artifacts...

Due to its relatively few assumptions, independent component analysis (ICA) has become a widely-used tool for the of functional magnetic resonance imaging (fMRI) data. In application, Infomax, been by far most frequently used ICA algorithm, primarily because it is first algorithm applied fMRI analysis. However, now there are number more flexible algorithms, which can exploit multiple types statistical properties signals with fewer assumptions. this work, we investigate performance Infomax...

Pain is subjective, while pain neuroimaging analysis methods need to be as objective possible. Most studies use a general linear model (GLM) approach which heavily relies on number of key assumptions. Independent component (ICA), the other hand, data-driven and hence significantly reduces for specific assumptions such Gaussian distributed residuals definition user-specified design matrix, both required by GLM. In this paper, we propose two-level ICA-based method an attractive alternative,...

Recently, much attention has been devoted to examining time-varying changes in functional connectivity understand the network structure human brain. Most studies, however, analyze but ignore spatial information. In this paper, we propose a method based on independent vector analysis (IVA) study dynamic (dFNC) as well (dsFNC) fMRI data. Though IVA allows one effectively capture both, its performance degrades with increase number of datasets. Hence, an effective scheme bypass limitation...

Functional magnetic resonance imaging (fMRI) is a powerful, noninvasive tool that has significantly contributed to the understanding of human brain. FMRI data provide sequence whole-brain volumes over time and hence are inherently four dimensional (4D). Missing in fMRI experiments arise from image acquisition limits, susceptibility motion artifacts or during confounding noise removal. Hence, significant brain regions may be excluded data, which can seriously undermine quality subsequent...