Abstract
The ability to accurately estimate effective connectivity among brain regions from neuroimaging data could help answering many open questions in neuroscience. We propose a method which uses causality to obtain a measure of effective connectivity from fMRI data. The method uses a vector autoregressive model for the latent variables describing neuronal activity in combination with a linear observation model based on a convolution with a hemodynamic response function. Due to the employed modeling, it is possible to efficiently estimate all latent variables of the model using a variational Bayesian inference algorithm. The computational efficiency of the method enables us to apply it to large scale problems with high sampling rates and several hundred regions of interest. We use a comprehensive empirical evaluation with synthetic and real fMRI data to evaluate the performance of our method under various conditions.
Highlights
Functional neuroimaging has been used to obtain spatial maps of brain activation, e.g., using functional magnetic resonance imaging or positron emission tomography (PET), or to study the spatio-temporal progression of activity using magneto- or electroencephalography (M/EEG)
In dynamic causal modeling (DCM), the relationship between neuronal activity in different regions of interest (ROIs) is described by bilinear ordinary differential equations (ODEs) and the functional magnetic resonance imaging (fMRI) observation process is modeled by a biophysical model based on the Balloon model (Buxton et al, 1998, 2004)
DCM is typically used for small numbers of ROIs and DCM methods typically are confirmatory approaches, i.e., the user provides a number of different candidate models describing the connectivity, which are ranked based on an approximation to the model evidence
Summary
Functional neuroimaging has been used to obtain spatial maps of brain activation, e.g., using functional magnetic resonance imaging (fMRI) or positron emission tomography (PET), or to study the spatio-temporal progression of activity using magneto- or electroencephalography (M/EEG). As there is variability in the shape of the HRF among brain regions and individuals (Handwerker et al, 2004) and the sampling rate of the MRI scanner is low, detecting effective connectivity from causal interactions that exist in the observed fMRI data is a challenging problem. We propose a causal connectivity method for fMRI which employs a VAR model of arbitrary order for the time series of neuronal activity in combination with a linear hemodynamic convolution model for the fMRI observation process. The results show that the proposed method offers some benefits over WGC, especially in low SNR situations and when HRF variations are present Both the proposed method and WGC can at times detect a causal influence with the opposite direction of the true influence, which is a known problem for WGC methods (David et al, 2008; Deshpande et al, 2010; Seth www.frontiersin.org et al, 2013). We apply the method to real fMRI data and conclude the paper
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
