A qualitative test of the balloon model for BOLD‐based MR signal changes at 3T

Abstract

The aim of this study was to adapt the balloon model for BOLDbased MR signal changes to a magnetic field strength of 3T and to examine its validity. The simultaneous measurement of BOLD and diffusion-weighted BOLD responses was performed. The amplitude of the BOLD peak was found to be similar for all subjects when a short visual stimulus of 6 sec was used. The rise-time to the BOLD peak and the shape and depth of the poststimulus undershoot varied significantly. A fit of the experimental BOLD responses was found to be possible by use of parameters within a reasonable physiological range. The relations between these parameters and their influence on the modeled BOLD responses is discussed. A prediction of the balloon model is the occurrence of a BOLD overshoot, i.e., a lag between the changes of the blood volume and the blood flow after the start of the stimulation. Functional MRI is based on measuring the local changes in the blood oxygenation level dependent (BOLD) contrast that occur as a consequence of neuronal activation in the brain (1-3). The change in the MR signal is determined by alterations in the concentration of paramagnetic deoxyhemoglobin, which modifies the effective transverse relaxation time T* 2 . Increases in oxygen consumption and blood volume lead to an increase in the deoxyhemoglobin content and hence to a decreased MR signal, whereas increases in blood flow have the opposite effect (4). Although considerable progress has been made in measuring the changes in oxygen consumption (5-7), blood volume (8,9), and cerebral blood flow (6,10,11), the transient response in the BOLD signal is still not completely understood. It can be divided into three intervals: the fast response (12-15), during which the signal change is negative, the period of elevated signal, and the poststimulus undershoot (3). In order to understand the origins of these signal changes it is desirable to relate the BOLD signal directly to the relevant physiological parameters.The most prominent approach developed to date for modeling the BOLD signal under nonsteady-state conditions is given by the balloon model of Buxton et al. (10). Based on the experimental work of Mandeville et al. (8), who showed that changes in blood volume occur subsequent to changes in blood flow, it considers the vasculature contributing to the BOLD signal as a swelling balloon. The basis of the model is a differential equation system for the blood volume and the total deoxyhemoglobin content, combined with an expression for the total signal change. The results of simulations for the intra-and extravascular signal contributions (16,17) and the oxygen limitation model (18) were used in order to derive time courses for the BOLD signal response.In order to study the time courses of the deoxyhemoglobin concentration and the blood volume in detail it is advantageous if the contributions of the intra-and extravascular spaces can be evaluated separately. This may be achieved by using a weak diffusion weighting to suppress the signal from the intravascular comp...

Investigating the post-stimulus undershoot of the BOLD signal—a simultaneous fMRI and fNIRS study

“…Therefore, it may be concluded that the poststimulus period had a higher inter-individual variability than the stimulus period in agreement with Mildner et al (2001).…”

“…All previous studies that combined fNIRS with another imaging method did not investigate the post-stimulus period of the hemodynamic response. To include that period, we chose a long intertrial interval of 60 s (Mildner et al, 2001). An event-related approach was employed to detect the hemodynamic response to a single trial (Burock et al, 1998;Schroeter et al, 2002Schroeter et al, , 2004c.…”

Modeling the hemodynamic response to brain activation

“…A similar windkessel model was proposed by Mandeville et al (1999b) to embody the same concept and provide a biomechanical mechanism for a delayed CBV return to baseline. The balloon model has been refined and compared with experimental data (Feng et al, 2001;Friston, 2002;Friston et al, 2000;Mildner et al, 2001;Obata et al, 2004;Toronov et al, 2003), and some errors in the original parameter estimates were recently corrected (Obata et al, 2004). The model is capable of producing BOLD poststimulus undershoots that match well with experimental data.…”

Independent components in stimulus-related BOLD signals and estimation of the underlying neural responses

“…The Friston-Buxton hemodynamic model involves nominally 11 parameters (see Appendix A). When fitting these, some authors (Mildner et al, 2001;Obata et al, 2004) have chosen to fix the values of some parameters (e.g., a, E 0 , V 0 , s 0 , a 1 and a 2 ) on the grounds that these parameters are known approximately from experiments. Others (Friston et al, 2000) have chosen to search on all parameters in light of experimental uncertainties and model assumptions, e.g., the widely cited value a c 0.38 relates f out and CBV only in steady-state (Grubb, 1974) and cannot be assumed to hold generally.…”