BOLD signal compartmentalization based on the apparent diffusion coefficient

Abstract

Functional MRI (fMRI) can detect blood oxygenation level dependent (BOLD) hemodynamic responses secondary to neuronal activity. The most commonly used method for detecting fMRI signals is the gradient-echo echo-planar imaging (EPI) technique because of its sensitivity and speed. However, it is generally believed that a significant portion of these signals arises from large veins, with additional contribution from the capillaries and parenchyma. Early experiments using diffusion-weighted gradient-echo EPI have suggested that intra-voxel incoherent motion (IVIM) weighting inherent in the sequence can selectively attenuate contributions from different vessels based on the differences in the mobility of the blood within them. In the present study, we used similar approach to characterize the apparent diffusion coefficient (ADC) distribution within the activated areas of BOLD contrast. It is shown that the voxel values of the ADCs obtained from this technique can infer various vascular contributions to the BOLD signal.

Complimentary aspects of diffusion imaging and fMRI: II. Elucidating contributions to the fMRI signal with diffusion sensitization

“…Several years after their original work, and at the considerably higher field strength of 4 T, Song et al [18][19][20][21][22][23][24] began a series of investigations comparing the fMRI signals generated by bBOLDQ contrast vs. bADCQ contrast. The general experimental arrangement was similar to the previous approach in which GE-EPI images were collected with cyclically varying b-factors, typically three.…”

“…In addition, the use of diffusion sensitization gradients [11], with or without velocity encoding or flow compensating schemes [12][13][14][15], can influence both the magnitude and spatial extent of the fMRI signal changes. Studies employing diffusion weighting methodology have been used to help identify and gauge the various contributions, both intra-and extravascular, to fMRI-based signal changes from water pools with different mobilities [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32]. In complimentary fashion, diffusion sensitization directly embedded within fMRI acquisitions [18,22,24,28], or diffusion data acquired separately and used for post-processing fMRI data [25], has proven of interest in enhancing the accuracy with which true neural activity might be localized.…”

Component structure of event-related fMRI responses in the different neurovascular compartments

“…Isotropic weighting gradients were embedded into a gradient-recalled spiral imaging sequence, as described in an earlier study [7]. For each 3-s cycle of the functional run, an ADC value was determined using a simple exponential fit based on the relationship that…”

“…The observed task-induced changes in the ADC primarily come from smaller vessels having a faster and more turbulent flow: arteries, arterioles and capillaries. The method of calculating ADCs applied in the present report, developed in recent studies by Song et al [7,8], uses a cyclical ramping of the diffusion weightings across sets of three consecutive time points and extracts the ADC by calculating an exponential fit to the signal across each of these time-point triplets. In addition, in this approach, the first time point of each three time points has a diffusion weighting of zero, thereby providing a conventional BOLD image for that time point.…”

“…Acquiring the BOLD images simultaneously with the ADC signals in this way provides two (orthogonal) contrasts with the BOLD mainly sensitive to the capillary and venous compartments and the ADC mainly sensitive to the arterial and capillary compartments. Combining the information from these different signal contrasts allows for the delineation of these various neurovascular compartments, with the intersection of these contrasts (i.e., those regions activated by both contrasts) corresponding to the capillary regions closely associated with the neuronal activities [7,8].…”