Academic Journal
Προβολή σε EDS
Συνδεδεμένο Πλήρες Κείμενο

Advancing temporal dynamics in spatial Random Field Theory: A framework for fMRI signal detection with simulation studies and task-based fMRI application.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Advancing temporal dynamics in spatial Random Field Theory: A framework for fMRI signal detection with simulation studies and task-based fMRI application.
Συγγραφείς: Acquah TBK; Department of Applied Statistics and Research Methods, University of Northern Colorado, Greeley, CO 80639, USA. Electronic address: acqu8971@bears.unco.edu., Shafie K; Department of Applied Statistics and Research Methods, University of Northern Colorado, Greeley, CO 80639, USA.
Πηγή: Journal of neuroscience methods [J Neurosci Methods] 2026 Aug; Vol. 432, pp. 110772. Date of Electronic Publication: 2026 Apr 21.
Τύπος έκδοσης: Journal Article
Γλώσσα: English
Στοιχεία περιοδικού: Publisher: Elsevier/North-Holland Biomedical Press Country of Publication: Netherlands NLM ID: 7905558 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1872-678X (Electronic) Linking ISSN: 01650270 NLM ISO Abbreviation: J Neurosci Methods Subsets: MEDLINE
Imprint Name(s): Original Publication: Amsterdam, Elsevier/North-Holland Biomedical Press.
Ιατρικοί όροι (MeSH): Magnetic Resonance Imaging*/methods , Brain Mapping*/methods , Brain*/physiology , Brain*/blood supply , Models, Neurological*, Image Processing, Computer-Assisted/methods ; Oxygen/blood ; Humans ; Computer Simulation ; Time Factors ; Algorithms
Περίληψη: Background: Functional magnetic resonance imaging (fMRI) data exhibit complex spatial and temporal dependencies, yet most statistical approaches rely on static or temporally independent models, limiting sensitivity to dynamic neural activity. This limitation is particularly critical in task-based studies, where neural responses evolve over time and across distributed brain networks.
New Method: We propose a time-adaptive random field framework that incorporates temporal dependence through a functional autoregressive process (FAR(1)), leading to a novel statistic, Xmax, for spatiotemporal signal detection.
Results: Simulation studies show that Xmax achieves substantially higher detection power than Ymax, with power exceeding 0.85 at high signal amplitude compared to below 0.65 for competing methods, while maintaining stable Type I error control. Application to real fMRI data demonstrates that Xmax produces more spatially extensive and coherent activation patterns across motor, language, visual, and subcortical regions, whereas GLM yields more localized and fragmented activations. These findings indicate improved sensitivity to temporally persistent and distributed neural signals.
Comparison With Existing Methods: Compared to Location-Scale random field theory (RFT) statistic Ymax and the classical GLM, which rely on static or temporally independent modeling, the proposed approach achieves improved power and produces more spatially coherent activation patterns by explicitly incorporating temporal dependence.
Conclusions: Incorporating temporal dependence improves sensitivity to dynamic neural signals and enables more effective detection of spatially distributed brain activity, providing a powerful framework for spatiotemporal analysis in neuroimaging.
(Copyright © 2026 Elsevier B.V. All rights reserved.)
Competing Interests: Declaration of competing interest The authors declare no competing interests. This research received no specific funding from public, commercial, or not-for-profit agencies.
Contributed Indexing: Keywords: Functional autoregressive model; Neural activation; Random Field Theory; Repeated measures ANOVA; Signal detection; Spatial–temporal modeling; Task-fMRI; fMRI
Substance Nomenclature: S88TT14065 (Oxygen)
Entry Date(s): Date Created: 20260423 Date Completed: 20260715 Latest Revision: 20260715
Update Code: 20260716
DOI: 10.1016/j.jneumeth.2026.110772
PMID: 42025970
Βάση Δεδομένων: MEDLINE
Περιγραφή
ISSN:1872-678X
DOI:10.1016/j.jneumeth.2026.110772