Snowball ICA: A Model Order Free Independent Component Analysis Strategy for Functional Magnetic Resonance Imaging Data

Hu, Guoqiang; Waters, Abigail B.; Aslan, Serdar; Frederick, Blaise; Cong, Fengyu; Nickerson, Lisa D.

doi:10.3389/fnins.2020.569657

Cited by 19 publications

(19 citation statements)

References 65 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To demonstrate that the SMN is contained within the group PCA space at low dimensionality, a snowball ICA was performed ( Hu et al, 2020 ). At each iteration, the most stable IC was subtracted from all subjects’ data.…”

Section: Resultsmentioning

confidence: 99%

“…To demonstrate that a canonical ICN is contained within the group PCA space, but absent from ICA results at a specific model order, a Snowball ICA ( Hu et al, 2020 ) was performed. In this procedure, at each iteration, the most stable IC was identified using bootstrap stability index I q , then respective back-reconstructed spatial maps were subtracted from subjects’ data.…”

Section: Methodsmentioning

confidence: 99%

“…Subsequent iterations then identified and removed the most stable IC from the data and repeated ICA. This procedure reliably extracts more ICNs than are initially present in ICA results for a single model order ( Hu et al, 2020 ).…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Stable Meta-Networks, Noise, and Artifacts in the Human Connectome: Low- to High-Dimensional Independent Components Analysis as a Hierarchy of Intrinsic Connectivity Networks

et al. 2021

View full text Add to dashboard Cite

Connectivity within the human connectome occurs between multiple neuronal systems—at small to very large spatial scales. Independent component analysis (ICA) is potentially a powerful tool to facilitate multi-scale analyses. However, ICA has yet to be fully evaluated at very low (10 or fewer) and ultra-high dimensionalities (200 or greater). The current investigation used data from the Human Connectome Project (HCP) to determine the following: (1) if larger networks, or meta-networks, are present at low dimensionality, (2) if nuisance sources increase with dimensionality, and (3) if ICA is prone to overfitting. Using bootstrap ICA, results suggested that, at very low dimensionality, ICA spatial maps consisted of Visual/Attention and Default/Control meta-networks. At fewer than 10 components, well-known networks such as the Somatomotor Network were absent from results. At high dimensionality, nuisance sources were present even in denoised high-quality data but were identifiable by correlation with tissue probability maps. Artifactual overfitting occurred to a minor degree at high dimensionalities. Basic summary statistics on spatial maps (maximum cluster size, maximum component weight, and average weight outside of maximum cluster) quickly and easily separated artifacts from gray matter sources. Lastly, by using weighted averages of bootstrap stability, even ultra-high dimensional ICA resulted in highly reproducible spatial maps. These results demonstrate how ICA can be applied in multi-scale analyses, reliably and accurately reproducing the hierarchy of meta-networks, large-scale networks, and subnetworks, thereby characterizing cortical connectivity across multiple spatial scales.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Stable Meta-Networks, Noise, and Artifacts in the Human Connectome: Low- to High-Dimensional Independent Components Analysis as a Hierarchy of Intrinsic Connectivity Networks

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Limited with selected model order, the ICA template is still not pure, which makes the voxel-level spatial distribution hard to be thresholded. Our previous study (Hu et al, 2020) has proposed an effective strategy, Snowball ICA, to address the dimensionality selection issue. In the further study, a more comprehensive template worth being further investigated with Snowball ICA.…”

Section: Discussionmentioning

confidence: 99%

Discovering hidden brain network responses to naturalistic stimuli via tensor component analysis of multi-subject fMRI data

Zhao

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The study of brain network interactions during naturalistic stimuli facilitates a deeper understanding of human brain function. Intersubject correlation (ISC) analysis of functional magnetic resonance imaging (fMRI) data is a widely used method that can measure neural responses to naturalistic stimuli that are consistent across subjects. However, interdependent correlation values in ISC artificially inflated the degrees of freedom, which hinders the investigation of individual differences. Besides, the existing ISC model mainly focus on similarities between subjects but fails to distinguish neural responses to different stimuli features. To estimate large-scale brain networks evoked with naturalistic stimuli, we propose a novel analytic framework to characterize shared spatio-temporal patterns across subjects in a purely data-driven manner. In the framework, a third-order tensor is constructed from the timeseries extracted from all brain regions from a given parcellation, for all participants, with modes of the tensor corresponding to spatial distribution, time series and participants. Tensor component analysis (TCA) will then reveal spatially and temporally shared components, i.e., naturalistic stimuli evoked networks, their temporal courses of activity and subject loadings of each component. To enhance the reproducibility of the estimation with TCA, a novel spectral clustering method, tensor spectral clustering, was proposed and applied to evaluate the stability of TCA algorithm. We demonstrate the effectiveness of the proposed framework via simulations and real fMRI data collected during a motor task with a traditional fMRI study design. We also apply the proposed framework to fMRI data collected during passive movie watching to illustrate how reproducible brain networks are identified evoked by naturalistic movie viewing.

show abstract

“…In addition, multi-modal order methods based on ICA have been proposed to deal with the selection of the optimal number of independent components in neuroimaging studies. For instance, [24] used healthy unrelated subjects from the WU-Minn Human Connectome Project [25] to show the improvement in the selection of components when the method follows a free order model. [26] also proposed a method to capture information at multiple model orders, showing an improvement in classifying brain patterns of schizophrenia individuals vs healthy controls.…”

Section: Methods For the Analysis Of Multiple Phenotypesmentioning

confidence: 99%

Multivariate Analysis and Modelling of multiple Brain endOphenotypes: Let’s MAMBO!

Vilor‐Tejedor

Garrido-Martín

Rodríguez‐Fernández

et al. 2021

Computational and Structural Biotechnology Journal

View full text Add to dashboard Cite

Imaging genetic studies aim to test how genetic information influences brain structure and function by combining neuroimaging-based brain features and genetic data from the same individual. Most studies focus on individual correlation and association tests between genetic variants and a single measurement of the brain. Despite the great success of univariate approaches, given the capacity of neuroimaging methods to provide a multiplicity of cerebral phenotypes, the development and application of multivariate methods become crucial. In this article, we review novel methods and strategies focused on the analysis of multiple phenotypes and genetic data. We also discuss relevant aspects of multi-trait modelling in the context of neuroimaging data.

show abstract

Snowball ICA: A Model Order Free Independent Component Analysis Strategy for Functional Magnetic Resonance Imaging Data

Cited by 19 publications

References 65 publications

Stable Meta-Networks, Noise, and Artifacts in the Human Connectome: Low- to High-Dimensional Independent Components Analysis as a Hierarchy of Intrinsic Connectivity Networks

Stable Meta-Networks, Noise, and Artifacts in the Human Connectome: Low- to High-Dimensional Independent Components Analysis as a Hierarchy of Intrinsic Connectivity Networks

Discovering hidden brain network responses to naturalistic stimuli via tensor component analysis of multi-subject fMRI data

Multivariate Analysis and Modelling of multiple Brain endOphenotypes: Let’s MAMBO!

Contact Info

Product

Resources

About