Decoupling of the hemodynamic and activation-induced delays in functional magnetic resonance imaging

Purpose: The objective of this study was to develop a technique to decouple the hemodynamic delay from the task-induced delay on a voxel-by-voxel basis in functional magnetic resonance imaging (fMRI) data sets. Prior fMRI studies have reported variability in temporal delays of up to several seconds among activated voxels. It is currently assumed that this variability primarily arises from differences in the onset of task-induced activation, although the precise source of these delays has not been well characterized. Here, we hypothesize that the total delay during task activation can be modeled as a combination of neuronal (caused by differences in onset of neuronal firing), vasomotor (caused by flow changes during activation), and transit (caused by differences in the red blood cell arrival time) delays. Method: Subjects were scanned using a sequential dynamic susceptibility contrast (DSC) protocol during rest and fMRI of the motor cortex using a bilateral finger-tapping task. The total delay was determined using correlation coefficient analysis, whereas the intrinsic delay was determined from the DSC MRI. Subtraction of the transit delay from the corresponding total delay for each voxel yielded the task-induced delay. Results: In all subjects, a transit delay of 2.3 (±1.1) seconds and a task-induced delay of 0.7 (±0.6) second was observed between voxels, which is in good agreement with reports in the literature using other techniques. These results demonstrate the feasibility of the DSC MRI for separating the hemodynamic and task-induced delays in fMRI studies. Conclusion: This approach has the potential to elucidate the temporal characteristics of the blood oxygenation level-dependent signal during fMRI as well as to further our understanding of the dynamics of the activation-induced signal in neuroimaging.