update_A_Chol

Update A for VB FC Bayesian brain mapping using Cholesky prior for FC

internal

Implements Bayesian brain mapping models, including the prior
ICA (independent components analysis) model proposed in Mejia et al. (2020)
<doi:10.1080/01621459.2019.1679638> and the spatial prior ICA model
proposed in proposed in Mejia et al. (2022)
<doi:10.1080/10618600.2022.2104289>. Both models estimate subject-level
brain as deviations from known population-level networks, which are
estimated using standard ICA algorithms. Both models employ an
expectation-maximization algorithm for estimation of the latent brain
networks and unknown model parameters. Includes direct support for 'CIFTI',
'GIFTI', and 'NIFTI' neuroimaging file formats.

Amanda Mejia

BayesBrainMap

Estimate Brain Networks and Connectivity with Population-Derived
Priors

Damon Pham

Daniel Spencer

Mary Beth Nebel

update_A_Chol function

<dl><dt>mu_tau2, mu_S, E_SSt</dt>
<dd>Most recent estimates of posterior
moments for these variables.</dd>
<dt>BOLD</dt>
<dd>(\(T \times V\) matrix) preprocessed fMRI data.</dd>
<dt>ntime, nNets, nvox</dt>
<dd>Number of timepoints, number of networks, number of locations</dd>
<dt>prior_FC</dt>
<dd>IW parameters for FC prior</dd>
<dt>final</dt>
<dd>If TRUE, return cov_A. Default: <code>FALSE</code></dd>
<dt>exact</dt>
<dd>If FALSE, at intermediate steps (final = <code>FALSE</code>) will attempt to
approximate V_k as described in the appendix of Mejia et al. 2024. Setting
<code>exact = TRUE</code> may be helpful when using spatial ESS correction,
which can lead to a lower rate of usable prior samples for the approximation.</dd>
<dt>usePar</dt>
<dd>Parallelize the computation? Default: <code>FALSE</code>. Can be the
number of cores to use or <code>TRUE</code>, which will use the number available minus two.</dd>
<dt>verbose</dt>
<dd>If <code>TRUE</code>, display progress of algorithm. Default: <code>FALSE</code>.</dd></dl>

Arguments

Update A for VB FC Bayesian brain mapping using Cholesky prior for FC — update_A_Chol

<dl>

<dt>mu_tau2, mu_S, E_SSt</dt>
<dd>Most recent estimates of posterior
moments for these variables.</dd>


<dt>BOLD</dt>
<dd>(\(T \times V\) matrix) preprocessed fMRI data.</dd>


<dt>ntime, nNets, nvox</dt>
<dd>Number of timepoints, number of networks, number of locations</dd>


<dt>prior_FC</dt>
<dd>IW parameters for FC prior</dd>


<dt>final</dt>
<dd>If TRUE, return cov_A. Default: <code>FALSE</code></dd>


<dt>exact</dt>
<dd>If FALSE, at intermediate steps (final = <code>FALSE</code>) will attempt to
approximate V_k as described in the appendix of Mejia et al. 2024. Setting
<code>exact = TRUE</code> may be helpful when using spatial ESS correction,
which can lead to a lower rate of usable prior samples for the approximation.</dd>


<dt>usePar</dt>
<dd>Parallelize the computation? Default: <code>FALSE</code>. Can be the
number of cores to use or <code>TRUE</code>, which will use the number available minus two.</dd>


<dt>verbose</dt>
<dd>If <code>TRUE</code>, display progress of algorithm. Default: <code>FALSE</code>.</dd>

</dl>

update_A_Chol: Update A for VB FC Bayesian brain mapping using Cholesky prior for FC

Description

Usage

Value

Arguments