Comparison of EPISTAR and T sub 2 *-weighted gadolinium-enhanced perfusion imaging in patients with acute cerebral ischemia

Publisher Website
Google Scholar
Abstract
To compare echo-planar imaging with signal targeting and alternating radiofrequency (EPISTAR), an arterial spin-labeling technique, to a T2*-weighted gadolinium-enhanced (T2*-WGE) MR perfusion technique for the evaluation of acute cerebrovascular disease. Twenty-one EPISTAR and T2*-WGE perfusion studies were performed on 18 patients with the clinical diagnosis of acute stroke (12 men, 6 women, age range 34 to 89 years, mean age 68 years). For qualitative analysis, perfusion studies of both techniques were grouped into categories (hyperperfusion, normal perfusion, delayed perfusion, or absent perfusion) and compared with a Wilcoxon signed rank test. Quantitative analysis was performed using signal intensity measurements in a region of interest that was defined by diffusion-weighted imaging abnormalities. These signal intensity measurements were compared with a mirror region in the contralateral unaffected hemisphere. Signal intensity ratios (infarcted region versus the unaffected contralateral region) were calculated and compared using a paired t test. Qualitative analysis demonstrated agreement between the two techniques in 17 of 21 studies (hyperfusion, n = 3 patients; normal perfusion, n = 3; delayed perfusion, n = 4; and absent perfusion, n = 7). In four studies, the two techniques disagreed when EPISTAR demonstrated absent and T2*-WGE perfusion demonstrated delayed perfusion (p > 0.05). Quantitative analysis revealed a mean signal intensity ratio of 0.73 +/- 0.79 for the T2*-WGE perfusion technique and 0.69 +/- 0.68 for the EPISTAR technique (p > 0.05). The noninvasive EPISTAR technique can assess perfusion abnormalities similarly to the T2*-WGE perfusion technique and may provide a valuable alternative in the diagnosis of acute stroke patients. Differences between the two techniques can be explained by the applied inflow times in the EPISTAR technique.