Purpose: To test the feasibility of noninvasive global assessment of cerebral hemodynamic impairment with use of resting-state blood oxygenation level-dependent functional magnetic resonance (MR) imaging.
Materials and methods: In this institutional review board-approved study, five patients with chronic hypoperfusion without neurologic impairment and six patients with acute stroke underwent 10-minute resting-state functional MR imaging and dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging, which was considered the standard of reference. All patients gave informed consent. The temporal shift of low-frequency signal fluctuations in each voxel compared with the averaged whole brain or global mean signal at resting-state functional MR imaging and the delay in time to peak at dynamic susceptibility-weighted contrast-enhanced perfusion imaging were computed with voxel-wise analysis. The similarity of the temporal delay maps obtained with resting-state functional MR imaging and perfusion data, as well as the stability of the resting-state functional MR imaging measurement, were evaluated with the Dice similarity coefficient (DSC) and the two-tailed t test (random-effect analysis).
Results: The brain tissue with normal perfusion at dynamic susceptibility-weighted contrast-enhanced imaging showed no delay to global mean signal at resting-state functional MR imaging, whereas areas of abnormal perfusion with delayed time to peak (3.4 seconds ± 2.1) showed a delay at resting-state functional MR imaging that was similar to the time to peak at dynamic susceptibility-weighted contrast-enhanced perfusion imaging, both in spatial coverage (mean DSC, 0.57 ± 0.16) and tendency (t = 5.1, P < .001). Resting-state functional MR imaging measurements were highly stable (mean DSC, 0.83 ± 0.12).
Conclusion: Resting-state functional MR imaging temporal-shift analysis can noninvasively demonstrate the extent and degree of perfusion delay in patients with hypoperfusion both with and without neurologic deficit.
© RSNA, 2013.