Multimodal subspace independent vector analysis captures latent subspace structures in large multimodal neuroimaging studies DOI Creative Commons
Xinhui Li, Peter Kochunov, Tülay Adalı

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 17, 2023

A key challenge in neuroscience is to understand the structural and functional relationships of brain from high-dimensional, multimodal neuroimaging data. While conventional multivariate approaches often simplify statistical assumptions estimate one-dimensional independent sources shared across modalities, between true latent are likely more complex - dependence may exist within span one or dimensions. Here we present Multimodal Subspace Independent Vector Analysis (MSIVA), a methodology capture both joint unique vector multiple data modalities by defining cross-modal unimodal subspaces with variable In particular, MSIVA enables flexible estimation varying-size their one-to-one linkage corresponding modalities. As demonstrate, main benefit ability subject-level variability at voxel level subspaces, contrasting rigidity traditional methods that share same components subjects. We compared initialization baseline baseline, evaluated all three five candidate subspace structures on synthetic datasets. show successfully identified ground-truth datasets, while failed detect high-dimensional subspaces. then demonstrate better detected structure two large datasets including MRI (sMRI) (fMRI), baseline. From subsequent subspace-specific canonical correlation analysis, brain-phenotype prediction, voxelwise brain-age delta our findings suggest estimated optimal strongly associated various phenotype variables, age, sex, schizophrenia, lifestyle factors, cognitive functions. Further, modality- group-specific regions related measures such as age (e.g., cerebellum, precentral gyrus, cingulate gyrus sMRI; occipital lobe superior frontal fMRI), sex cerebellum sMRI, fMRI, precuneus sMRI schizophrenia temporal pole, operculum cortex lingual shedding light phenotypic neuropsychiatric biomarkers linked function.

Language: Английский

Dev-Atlas: A reference atlas of functional brain networks for typically developing adolescents DOI Creative Commons
Gaëlle E. Doucet,

Callum Goldsmith,

K. Myers

et al.

Developmental Cognitive Neuroscience, Journal Year: 2025, Volume and Issue: 72, P. 101523 - 101523

Published: Feb. 8, 2025

It is well accepted that the brain functionally organized into multiple networks and extensive literature has demonstrated organization of these shows major changes during adolescence. Yet, there limited option for a reference functional atlas derived from typically-developing adolescents, which problematic as reliable identification crucially depends on use such atlases. In this context, we utilized resting-state MRI data 1391 youth aged 8-17 years to create an adolescent-specific networks. We further investigated impact age sex Using multiscale individual component clustering algorithm, identified 24 networks, classified within six domains: Default-Mode (5 networks), Control (4 Salience (3 Attention Somatomotor Visual networks). large effects spatial topography majority network connectivity. Sex were not widespread. created novel atlas, named Dev-Atlas, focused sample, with hope can be used in future developmental neuroscience studies.

Language: Английский

Citations

0
Building Multivariate Molecular Imaging Brain Atlases Using the NeuroMark PET Independent Component Analysis Framework DOI Creative Commons
Cyrus Eierud, Martin Nørgaard, Murat Bilgel

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 23, 2025

Abstract Molecular imaging analyses using positron emission tomography (PET) data often rely on macro-anatomical regions of interest (ROI), which may not align with chemo-architectural boundaries and obscure functional distinctions. While methods such as independent component analysis (ICA) have been useful to address this limitation, the fully data-driven nature can make it challenging compare results across studies. Here, we introduce NeuroMark PET approach, utilizing spatially constrained define overlapping that reflect brain’s molecular architecture. We first generate an ICA template for radiotracer florbetapir (FBP), targeting amyloid-β (Aβ) accumulation in brain, blind large datasets identify replicable components. Only components targeted Aβ were included study, defined networks (AβNs), by omitting myelin or other non-Aβ targets. Next, use AβNs priors ICA, resulting a automated pipeline called PET. This pipeline, including its AβNs, was validated against standard neuroanatomical atlas, from Alzheimer’s Disease Neuroimaging Initiative (ADNI). The study 296 cognitively normal participants FBP scans 173 florbetaben (FBB) scans, analogue also accumulation. Our show captures biologically meaningful, participant-specific features, subject specific loading values, consistent individuals, shows higher sensitivity power detecting age-related changes compared traditional atlas-based ROIs. Using framework, highlight some advantages data. In AβN consists weighted voxels forms pattern throughout entire brain. For example, values at every voxel overlap one another, enabling separation artifacts coincide interest. addition, approach allows differentiation, separating white matter components, complex ways mainly residing neighboring gray matter. Results showed most age associated (representing cognitive control network, CC1) exhibited stronger association suggest each represents spatial network following uptake greater biological relevance anatomical summary, proposed offers providing accurate reproducible brain AβNs. enhances our ability investigate underpinnings function pathology, offering alternative ROI-based analyses.

Language: Английский

Citations

0
Cognitive and psychiatric relevance of dynamic functional connectivity states in a large (N > 10,000) children population DOI Creative Commons
Zening Fu, Jing Sui, Armin Iraji

et al.

Molecular Psychiatry, Journal Year: 2024, Volume and Issue: unknown

Published: July 31, 2024

Children's brains dynamically adapt to the stimuli from internal state and external environment, allowing for changes in cognitive mental behavior. In this work, we performed a large-scale analysis of dynamic functional connectivity (DFC) children aged 9~11 years, investigating how brain dynamics relate performance health at an early age. A hybrid independent component framework was applied Adolescent Brain Cognitive Development (ABCD) data containing 10,988 children. We combined sliding-window approach with k-means clustering identify five states distinct DFC patterns. Interestingly, occurrence strongly connected most within-network synchrony anticorrelations between networks, especially sensory networks cerebellum other negatively correlated positively dimensional psychopathology Meanwhile, opposite relationships were observed showing integration antagonism default-mode sensorimotor but weak segregation cerebellum. The mediation further showed that attention problems mediated effect on performance. This investigation unveils neurological underpinnings states, which suggests tracking transient may help characterize guide people provide intervention buffer adverse influences.

Language: Английский

Citations

3
Studying time-resolved functional connectivity via communication theory: on the complementary nature of phase synchronization and sliding window Pearson correlation. DOI Creative Commons
Sir-Lord Wiafe,

Nana O. Asante,

Vince D. Calhoun

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: June 13, 2024

Abstract Time-resolved functional connectivity (trFC) assesses the time-resolved coupling between brain regions using magnetic resonance imaging (fMRI) data. This study aims to compare two techniques used estimate trFC, investigate their similarities and differences when applied fMRI These are sliding window Pearson correlation (SWPC), an amplitude-based approach, phase synchronization (PS), a phase-based technique. To accomplish our objective, we resting-state data from Human Connectome Project (HCP) with 827 subjects (repetition time: 0.7s) Function Biomedical Informatics Research Network (fBIRN) 311 2s), which included 151 schizophrenia patients 160 controls. Our simulations reveal distinct strengths in methods: SWPC captures high-magnitude, low-frequency connectivity, while PS detects low-magnitude, high-frequency connectivity. Stronger correlations align pronounced oscillations. For data, higher occur matched frequencies smaller sizes (∼30s), but larger windows (∼88s) sacrifice clinically relevant information. Both methods identify schizophrenia-associated network state show different patterns: highlights low anti-correlations visual, subcortical, auditory, sensory-motor networks, shows reduced positive among these networks. In sum, findings underscore complementary nature of PS, elucidating respective limitations without implying superiority one over other.

Language: Английский

Citations

2
Born Connected: Do Infants Already Have Adult-Like Multi-Scale Connectivity Networks? DOI Creative Commons

Prerana Bajracharya,

Shiva Mirzaeian,

Zening Fu

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 27, 2024

Abstract The human brain undergoes remarkable development with the first six postnatal months witnessing most dramatic structural and functional changes, making this period critical for in-depth research. rsfMRI studies have identified intrinsic connectivity networks (ICNs), including default mode network, in infants. Although early formation of these has been suggested, specific identification number ICNs reported infants vary widely, leading to inconclusive findings. In adults, provided valuable insights into function, spanning various mental states disorders. A recent study analyzed data from over 100,000 subjects generated a template 105 multi-scale enhancing replicability generalizability across studies. Yet, presence not investigated. This addresses significant gap by evaluating infants, offering insight stages establishing baseline longitudinal To accomplish goal, we employ two sets analyses. First, fully data-driven approach investigate infant itself. Towards aim, also introduce burst independent component analysis (bICA), which provides reliable unbiased network identification. results reveal showing high correlation (rho > 0.5), highlighting potential continuity such We next demonstrate that reference-informed ICA-based techniques can reliably estimate feasibility leveraging NeuroMark framework robust extraction. only enhances cross-study comparisons lifespans but facilitates changes different age ranges. summary, our highlights novel discovery already possesses are widely observed older cohorts.

Language: Английский

Citations

1
Dev-Atlas: A reference atlas of functional brain networks for typically developing adolescents DOI Creative Commons
Gaëlle E. Doucet,

Callum Goldsmith,

K. Myers

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 20, 2024

Abstract Adolescence is a critical period for neural changes, including maturation of the brain’s cognitive networks, but also increased vulnerability to psychopathology. It well accepted that brain functionally organized into multiple interacting networks and extensive literature has demonstrated spatial functional organization these shows major age-related changes across lifespan, particularly during adolescence. Yet, there limited option reference atlas derived from typically developing adolescents, which especially problematic as reliable reproducible identification crucially depends on use such atlases. In this context, we utilized resting-state MRI data total 1,391 youth between ages 8 17 years create new adolescent-specific networks. We further investigated impact age sex Using multiscale individual component clustering algorithm (MICCA), identified 24 classified within six domains: Default-Mode (5 networks), Control (4 Salience (3 Attention Somatomotor Visual networks). large effects topography majority network connectivity (FNC) The DMN showed reduced FNC with other older age. Sex were not widespread. No significant sex-by-age interactions detected. Overall, created novel atlas, named Dev-Atlas, focused sample, hope can be used in future independent developmental neuroscience studies. Dev-Atlas freely available research community.

Language: Английский

Citations

0
Spontaneous Brain Dynamics Associated With Acceleration Of Longterm Functional Connectome In Postnatal Development DOI Creative Commons

Liang Ma,

Sarah Shultz, Zening Fu

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 14, 2024

ABSTRACT The first six postnatal months are a critical period for brain development, marked by rapid changes in functional neural circuits. However, long-term neonatal connectome lacks an interpretive imaging indicator the future development due to non-linearity characteristics. In this study, we introduce approach extract intrinsic states from short-term dynamics study (longitudinal) development. We found high association (r=0.460) between co-activated pattern of specific state and acceleration non-linear static connectome. fractional occupancy, self-sustaining probability share similar age tendency with change rate within majority function These findings suggest that could serve as potential biomarkers predicting

Language: Английский

Citations

0
Multimodal subspace independent vector analysis captures latent subspace structures in large multimodal neuroimaging studies DOI Creative Commons
Xinhui Li, Peter Kochunov, Tülay Adalı

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 17, 2023

A key challenge in neuroscience is to understand the structural and functional relationships of brain from high-dimensional, multimodal neuroimaging data. While conventional multivariate approaches often simplify statistical assumptions estimate one-dimensional independent sources shared across modalities, between true latent are likely more complex - dependence may exist within span one or dimensions. Here we present Multimodal Subspace Independent Vector Analysis (MSIVA), a methodology capture both joint unique vector multiple data modalities by defining cross-modal unimodal subspaces with variable In particular, MSIVA enables flexible estimation varying-size their one-to-one linkage corresponding modalities. As demonstrate, main benefit ability subject-level variability at voxel level subspaces, contrasting rigidity traditional methods that share same components subjects. We compared initialization baseline baseline, evaluated all three five candidate subspace structures on synthetic datasets. show successfully identified ground-truth datasets, while failed detect high-dimensional subspaces. then demonstrate better detected structure two large datasets including MRI (sMRI) (fMRI), baseline. From subsequent subspace-specific canonical correlation analysis, brain-phenotype prediction, voxelwise brain-age delta our findings suggest estimated optimal strongly associated various phenotype variables, age, sex, schizophrenia, lifestyle factors, cognitive functions. Further, modality- group-specific regions related measures such as age (e.g., cerebellum, precentral gyrus, cingulate gyrus sMRI; occipital lobe superior frontal fMRI), sex cerebellum sMRI, fMRI, precuneus sMRI schizophrenia temporal pole, operculum cortex lingual shedding light phenotypic neuropsychiatric biomarkers linked function.

Language: Английский

Citations

1