Abstract:Previous work investigated a range of spatio-temporal constraints for fMRI data analysis to provide robust detection of neural activation. We present a mixture-based method for the spatio-temporal modelling of fMRI data. This approach assumes that fMRI time series are generated by a probabilistic superposition of a small set of spatio-temporal prototypes (mixture components). Each prototype comprises a temporal model that explains fMRI signals on a single voxel and the model's “region of influence” through a spatial prior over the voxel space. As the key ingredient of our temporal model, the Hidden Process Model (HPM) framework proposed in Hutchinson et al. (2009) is adopted to infer the overlapping cognitive processes triggered by stimuli. Unlike the original HPM framework, we use a parametric model of Haemodynamic Response Function (HRF) so that biological constraints are naturally incorporated in the HRF estimation. The spatial priors are defined in terms of a parameterised distribution. Thus, the total number of parameters in the model does not depend on the number of voxels. The resulting model provides a conceptually principled and computationally efficient approach to identify spatio-temporal patterns of neural activation from fMRI data, in contrast to most conventional approaches in the literature focusing on the detection of spatial patterns. We first verify the proposed model in a controlled experimental setting using synthetic data. The model is further validated on real fMRI data obtained from a rapid event-related visual recognition experiment (Mayhew et al., 2012). Our model enables us to evaluate in a principled manner the variability of neural activations within individual regions of interest (ROIs). The results strongly suggest that, compared with occipitotemporal regions, the frontal ones are less homogeneous, requiring two HPM prototypes per region. Despite the rapid event-related experimental design, the model is capable of disentangling the perceptual judgement and motor response processes that are both activated in the frontal ROIs. Spatio-temporal heterogeneity in the frontal regions seems to be associated with diverse dynamic localizations of the two hidden processes in different subregions of frontal ROIs.

A robust model for spatiotemporal dependencies

An efficient approach to characterize spatio-temporal dependence in cortical surface fMRI data

Functional MRI analysis by a novel spatiotemporal ICA algorithm

Bayesian Analysis of fMRI data with Spatially-Varying Autoregressive Orders

A spatial mixture approach to inferring sub-ROI spatio-temporal patterns from rapid event-related fMRI data.

Time-varying correlation structure estimation and local-feature detection for spatio-temporal data

Spatiotemporal Blind Source Separation Using Double-Sided Approximate Joint Diagonalization.

Spatio-temporally separable non-linear latent factor learning: an application to somatomotor cortex fMRI data

Undersampled Dynamic MRI Reconstruction by Double Sparse Spatiotemporal Dictionary.

Towards a predictive spatio-temporal representation of brain data

Hierarchical Spatio-Temporal State-Space Modeling for fMRI Analysis

A novel method for sparse dynamic functional connectivity analysis from resting-state fMRI

Sparse plus low-rank autoregressive identification in neuroimaging time series

Time Series Analysis of fMRI Data: Spatial Modelling and Bayesian Computation

Unified Principal Component Analysis for Sparse and Dense Functional Data under Spatial Dependency

A method to find temporal structure of neuronal coactivity patterns with across-trial correlations

Dynamic Multivariate Functional Data Modeling Via Sparse Subspace Learning

Spatial–temporal Modelling of Fmri Data Through Spatially Regularized Mixture of Hidden Process Models

Efficient Covariance Estimation from Temporal Data

Sparse network estimation for dynamical spatio-temporal array models

Asymmetric canonical correlation analysis of Riemannian and high-dimensional data