Abstract
The blood oxygen level dependent (BOLD) functional magnetic resonance imaging (fMRI) signal arises as a consequence of changes in blood flow (CBF) and oxygen usage (CMR$_{O_2}$) that in turn are modulated by changes in neuronal activity. Much attention has been given to both theoretical and experimental aspects of the energetics but not to the neuronal activity. Here we use our previous theory relating the steady-state BOLD signal to neuronal activity and amalgamate it with the standard dynamic causal model (DCM; Friston) theory to produce a quantitative model relating the time-dependent BOLD signal to the underlying neuronal activity. Unlike existing treatments, this new theory incorporates a non-zero baseline activity in a completely consistent way, and is thus able to account for both positive and negative BOLD signals. It can reproduce a wide variety of experimental BOLD signals reported in the literature solely by adjusting the neuronal input activity. In this way it provides support for the claim that the main features of the signals, including post-stimulus undershoot and overshoot, are principally a result of changes in neuronal activity.

This publication has 45 references indexed in Scilit: