On the sensitivity of ASL MRI in detecting regional differences in cerebral blood flow

Abstract

Arterial-spin-labeling (ASL) MRI provides a non-invasive tool to measure cerebral blood flow (CBF) and is increasingly used as a surrogate for baseline neural activity. However, the power of ASL MRI in detecting CBF differences between patient and control subjects is hampered by inter-subject variations in global CBF, which are associated with non-neural factors and may contribute to the noise in the across-group comparison. Here, we investigated the sensitivity of this technique and proposed a normalization strategy to better detect such a difference. A “model” situation was employed in which two visual stimuli (i.e. cross fixation and flashing checkerboard) were presented to two groups of subjects to mimic “control” and “patient” groups (N=7 for each group), respectively. It was found that absolute CBF (aCBF) in the occipital lobe in the checkerboard group was 26.0% greater compared to the fixation group, but the level of significance was modest (p=0.03). In contrast, when normalizing the CBF with whole-brain CBF or CBF in a reference region (termed relative CBF, rCBF), the statistical significance was improved considerably (p<0.003). For voxel based analysis, the rCBF indices correctly detected CBF differences in the occipital lobe in the across-group comparison, while aCBF failed to detect any significant cluster using the same statistical threshold. We also performed Monte Carlo simulation to confirm the experimental findings and found that the power improvement was most pronounced when SNR is moderate and the underlying CBF difference was small. The simulation also showed that, with the proposed normalization, a detection power of 80% can be achieved using a sample size of about 20. In summary, rCBF is a more sensitive index to detect small differences in CBF, rather than the much-sought-after aCBF, since it reduces data noise caused by inter-subject variations in global CBF.

Cerebral Blood Flow Improvement after Indirect Revascularization for Pediatric Moyamoya Disease: A Statistical Analysis of Arterial Spin-Labeling MRI

“…To minimize intersubject variations of CBF and to increase the sensitivity of this imaging technique, we used whole-brain normalization and voxel-based analysis of ASL MR imaging, as previously described in cognitive studies. 7 At the group level, we were able to display preoperative hypoperfusion in the MCA territory and a postoperative increase of CBF perfusion, therefore validating the surgical technique. At the individual level, FIG 5.…”

“…Because children under-going an operation for MM need repeated quantification of their CBF before and after surgery, ASL MR imaging may be of particular interest, and statistical approaches, as described with statistical parametric analysis (SPM) of ASL MR imaging in cognitive series, may allow intraindividual and group analysis of CBF comparisons before and after the operation. 7 This could allow a noninvasive follow-up of patients having undergone an operation and help to prevent recurrent stroke in patients with persistent hypoperfusion.…”

Estimating the sample size required to detect an arterial spin labelling magnetic resonance imaging perfusion abnormality in voxel-wise group analyses

“…Despite some limitations, this study adds to the literature on statistical inference in functional neuroimaging and the results were in a range established in previous sample-size estimations (Aslan and Lu, 2010;Murphy et al, 2011;Desmond and Glover, 2002;Murphy and Garavan, 2004). Sample size requirements determined using FEAT were comparable to previous work on sample-size estimation where BOLD MRI acquisition was used (Desmond and Glover, 2002;Murphy and Garavan, 2004).…”

“…A BOLD-based fMRI study for an unpaired design scenario reported that 20 subjects per group could be sufficient for 80% power in reproducing truly activated voxels (Murphy and Garavan, 2004). A study based on ASL data has reported similar numbers of 20 subjects per group for a moderate 10% group perfusion difference at 80% power (Aslan and Lu, 2010), or between 20 to 40 subjects per group from an ROI analysis in Murphy et al (Murphy et al, 2011). The higher sample size requirement in Randomise as compared to previous sample size studies may be consequence of the stringent permutation approach applied to control for multiple comparisons.…”

“…In a recent study, a Monte Carlo simulation was used to model a 10% perfusion difference between two groups and considering a single voxel in the brain (Aslan and Lu, 2010). By normalising CBF values, the authors show that an 80% true-positive detection rate can be achieved with 20 participants per group.…”