The effect of cerebral blood perfusion on the correlation between cerebral stroke onset time and synthetic T2 mapping: a pilot study.

Background: In patients with acute stroke with an unknown onset time, the T2 relaxation time (qT2) in the region of diffusion restriction is associated with the time from symptom onset. We hypothesized that cerebral blood flow (CBF) status assessed using arterial spin labeling magnetic resonance (MR) imaging would influence the association between qT2 and stroke onset time. The purpose of this study was to preliminarily investigate the effects of diffusion-weighted imaging-T2-weighted fluid-attenuated inversion recovery (DWI-T2-FLAIR) mismatch and T2 mapping value changes on the accuracy of stroke onset time in patients with different CBF perfusion statuses. Methods: A total of 94 patients with acute ischemic stroke (symptom onset time ≤24 h) admitted to the Liaoning Thrombus Treatment Center of Integrated Chinese and Western Medicine, Liaoning, China, were enrolled in this cross-sectional retrospective study. MR image compilation (MAGiC), DWI, 3-dimensional (3D) pseudo-continuous arterial spin labeling perfusion (pcASL), and T2-FLAIR images were acquired. The T2 map was directly generated from MAGiC. The CBF map was assessed using 3D pcASL. Patients were divided into the good CBF group (CBF >25 mL/100 g/min) and the poor CBF group (CBF ≤25 mL/100 g/min). The T2 relaxation time (qT2), T2 relaxation time ratio (qT2 ratio), and T2-FLAIR signal intensity ratio (T2-FLAIR ratio) between the ischemic and nonischemic region of the contralateral side were calculated. The correlations between the qT2, qT2 ratio, T2-FLAIR ratio, and stroke onset time were statistically analyzed in the different CBF groups. Results: In DWI-restricted areas, the time from symptom onset correlated with the qT2 and T2-FLAIR ratio. We identified an interaction between this association and CBF status. In the poor CBF group, stroke onset time most significantly correlated with the qT2 ratio (r=0.493; P<0.001), followed by the qT2 (r=0.409; P=0.001) and the T2-FLAIR ratio (r=0.385; P=0.003). In the total patient group, the stroke onset time moderately correlated with the qT2 ratio (r=0.438; P<0.001) but weakly correlated with the qT2 (r=0.314; P=0.002) and the T2-FLAIR ratio (r=0.352; P=0.001). In the good CBF group, no obvious correlations were found between stroke onset time and all MR quantitative indicators. Conclusions: In patients with reduced cerebral perfusion, the stroke onset time correlated with changes in the T2-FLAIR signal and qT2. The stratified analysis showed that the qT2 ratio had a higher correlation with stroke onset time than with the qT2 and T2-FLAIR ratio.