Abstract
Recently, MRI arterial spin labeling (ASL) is attracting attention as the noninvasive and convenient imaging technique to evaluate cerebral circulation. However, since there is a problem regarding to arterial transit time, it remains unclear that the reliability of this new modality is equivalent to the established one, such as SPECT or PET. In this study, we analyzed cerebral circulation for 20 cases with chronic ischemic cerebrovascular disease by multiple post-labeling delay (PLD) ASL and 15O-PET to determine if ASL can truly evaluate cerebral circulation. As the results, both CBF/CBV and CBF obtained by 15O-PET were significantly correlated with PLD (1000 and 1500 msec) by ASL. In addition, there was significantly negative correlation between CBV by 15O-PET and PLD (1000 and 1500 msec) by ASL. In the cases whose right-and-left difference of CBF was more than 10%, the receiver operating characteristic (ROC) analysis of revealed area under curve (AUC) of PLD (1000 msec) was as the highest as 0.987 (p=0.001) among all PLD. The cut off value of right-and-left difference by ASL PLD (1000 msec) was 31% in the cases whose right-and-left difference of CBF was more than 10%. The sensitivity and specificity of this analysis was 100% and 93.3%, respectively. In conclusion, we suggest that short PLD, especially PLD (1000 msec) reflects the cerebral perfusion pressure. Therefore, ASL could be useful for the screening of the cerebral circulation in cases of the chronic ischemic cerebrovascular disease.
