Modeling of human brain activation -induced signal changes in functional MRI

Brain functional magnetic resonance imaging (fMRI) is possible because of local changes in cerebral blood flow (CBF), cerebral blood volume (CBV) and oxygenation level. Understanding the role of involved mechanisms and parameters is important for interpreting fMRI results as well as for the design of experiments. In this study, we focused on two specific goals: one was to develop a methodology for determining the relative signal changes based on physiological parameters, and a related goal was to determine in a more quantitative way, how changes in CBV and cerebral metabolic rate of oxygen (CMRO2) affect fMRI signal changes.;Our numerical simulations were primarily designed to quantify the fMRI signal change and its dependency on changes in relaxation rate through magnetic susceptibility difference between intra- and extravascular space. The results show that the extravascular signal by itself inadequately describes fMRI signal changes at 1.5 T, so the study of combined intra- and extravascular contributions is necessary. Two main approaches have been proposed. One is based on maintaining constant blood volume. We have developed a mathematical model for determining percent signal change in fMR1 studies. The model enables investigating individual mechanism such as BOLD (blood oxygenation level dependent) or/and inflow effects. Results show a strong contribution to the net fMRI signal from inflow effect as well as BOLD mechanism.;The second approach is based on a model of the dynamic characteristics of CBF, CBV and deoxyhemoglobin changes for estimating the transient aspects of BOLD signal. The model predicts that in addition to CBF, other parameters such as CBV and CMRO2 are important to interpret the signal changes in T2*-weighted images. According to the results, a negative signal change and a corresponding increase in deoxyhemoglobin occurs when CMRO2 increases prior to blood flow at the onset of the stimulus. The results demonstrate a pronounced post-stimulus undershoot that is not present in the estimated BOLD signal change while the blood volume is maintained constant, and show that the post-stimulus undershoot occurs when blood volume returns to baseline more slowly than flow.