Causal Modelling and Brain Connectivity in Functional Magnetic Resonance Imaging (original) (raw)
Figure 2
Modelling Haemodynamics
(A) This schematic shows the architecture of a haemodynamic model for a single region. Neuronal activity induces a vasodilatory and activity-dependent signal s that increases blood flow f. Flow causes changes in volume and deoxyhemoglobin (v and q). These two haemodynamic states enter an output non-linearity to give the observed fMRI signal y. (B) This transformation from neuronal states x(t)i to haemodynamic response y(t)i is encoded graphically by the boxes in the previous figure and corresponds to a convolution. The implicit convolution kernel or haemodynamic response function is shown in the insert for typical values of the haemodynamic model's parameters.