Sigmoid function parameter stability in anatomically informed priors for dynamic causal models

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26220%2F14%3APU109412" target="_blank" >RIV/00216305:26220/14:PU109412 - isvavai.cz</a>
Result on the web
<a href="http://ww4.aievolution.com/hbm1401/index.cfm?do=abs.viewAbs&abs=3359" target="_blank" >http://ww4.aievolution.com/hbm1401/index.cfm?do=abs.viewAbs&abs=3359</a>
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Sigmoid function parameter stability in anatomically informed priors for dynamic causal models
Original language description
Stephan et al. (2009b) introduced and showed that the anatomical connectivity (AC) can improve effective connectivity estimation via dynamic causal modelling (DCM) (Friston et al. 2003). They proved it on the model with 4 regions of interest (ROIs) in visual cortex. But it can not be expected that the sigmoid parameters which affects the variance of shrinkage priors based on AC will be stable and same for different brain areas or especially for different number of ROIs. Additionally, usage of the fully-connected model could be more beneficial in this application. The aim of this work was comparison of optimal sigmoid function parameters in approach introduced by Stephan et al. (2009b) on 6 ROIs fully-connected model of motor task data with optimal parameters estimated by the authors in 2009.
Czech name
—
Czech description
—

Project
<a href="/en/project/GAP103%2F12%2F0552" target="_blank" >GAP103/12/0552: Comparison and inference of methods for evaluation of functional and effective connectivity in fMRI</a><br>
Continuities
S - Specificky vyzkum na vysokych skolach

Publication year
2014
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Similar results(10)