3D Pseudo Continuous Arterial Spin Labelling Perfusion MR Imaging In Patients With Hyperacute Stroke

Objective: Magnetic resonance imaging is an important imaging method for early diagnosis and guidance of acute ischemic stroke.MRI can not only provide clinical indications of the location and extent of acute ischemic stroke,but also provide brain tissue diffusion,perfusion,biochemistry and Information on metabolism,understanding the extent of ischemia,whether thrombolytic therapy,and whether the blood supply area survives and recovers after thrombolytic therapy,and then comprehensively evaluate acute ischemic stroke.Perfusion MRI is used to diagnose patients with acute ischemic stroke to guide clinical reperfusion therapy.However,renal-derived systemic fibrosis limits the use of contrast agents.The purpose of our study was to evaluate an alternative non-invasive perfusion technique---Three-dimentional pseudo-continuous arterial spin labeling(3D PC-ASL)technique Dynamic susceptibility contrast perfusion weighted imaging(DSC-PWI)was compared to evaluate the ability of 3D PC-ASL technique in detecting perfusion defects.Methods: A total of 35 patients with suspected hyperacute or acute ischemic stroke were enrolled in our hospital from June 2016 to June 2017.MRI was performed on routine T2 WI,T1WI,T2 FLAIR,and diffusion-weighted imaging.(DWI),3D PC-ASL and DSC-PWI examinations.The images obtained by 3D PC-ASL and DSC-PWI were compared and the images obtained were randomly assigned to two senior physicians for double-blind reading and diagnosis of image quality,presence of perfusion defect and diffusion-perfusion mismatch area.Chi-square test was used to determine whether 3D PC-ASL and DSC-PWI were consistent in detecting the success of perfusion defects and perfusion/diffusion mismatch.Kappa statistic was used to perform an inter-rater reliability analysis to determine consistency between evaluators.Results: From June 2016 to June 2017,35 patients with acute ischemic stroke were suspected,including 23 male patients and 12 female patients.The male to female ratio was approximately 2:1 and the age was 64.5±12 years old..The time from symptom onset to imaging was 5.5 ± 2.7 hours.Of the 35 patients,4 were excluded because they were unable to receive angiography due to GFR <30 m L/min.Thirty-one patients underwent 3D PC-ASL and DSC perfusion imaging,of which 71%(22/31)were clinically diagnosed as stroke.The image quality in DSC images was excellent,good,acceptable,and poor,respectively 22(71%),5(16%),2(6.5%),and 2(6.5%).The image quality in the 3D PC-ASL image was excellent(good,good,good,and poor)of 23(74.2%),6(19.4%),1(3.2%),and 1(3.2%),respectively.Of the 22 patients with confirmed stroke,14(63.6%)were hypoperfused with DSC.3D PC-ASL found 11(78.6%)hypoperfusion in 14 patients with insufficient DSC perfusion.Therefore,in the remaining 3 stroke patients,DSC showed insufficient perfusion,but 3D PC-ASL did not show hypoperfusion;2 patients had cortical gray matter loss and 1 patient had basal ganglion loss.Eight DSCs showed normal perfusion of stroke patients,three of whom showed hypoperfusion on 3D PC-ASL images.The 3 patients with 3D PC-ASL showed perfusion defects in the basal ganglia,while DSC showed no perfusion defects.Chi-square test was used to evaluate the results of perfusion defects in the images of stroke patients by two methods(X2=2.15,P>0.05).Kappa test was used to evaluate the consistency of the perfused group between the two physicians using DSC and 3D PC-ASL imaging,with kappa values of 0.942 and 0.881,respectively.Of the 22 patients with confirmed stroke,8(31%)had significant perfusion/diffusion mismatched areas on DSC,and 6(90%)had perfusion/diffusion mismatched areas on 3D PC-ASL.There were perfusion/diffusion mismatched areas on the DSC,but two patients with no perfusion/diffusion mismatched areas on the 3D PC-ASL had cortical lesions.Of the 14 patients with insignificant DSC perfusion/diffusion mismatch,4(33%)showed significant perfusion/diffusion mismatched areas on 3D PC-ASL(Table 3).Chi-square test was used to evaluate the results of two methods for perfusion/diffusion mismatch in the images of stroke patients(X2=2.75,P>0.05).Kappa test was used to evaluate the inter-group consistency of the perfusion/diffusion mismatched areas by DSC and 3D PC-ASL imaging,respectively,with kappa values of 0.815 and 0.765,respectively.Two of the 22 patients with clinically diagnosed stroke(9%)were unable to measure the volume of the perfusion defect because of the poor quality of the DSC image.The volume of DSC hypoperfusion in stroke patients was 143.8 ± 95.3 ml.In patients with no perfusion defects on 3D PC-ASL,the volume of hypoperfusion on the DSC was 53 ± 42.6 ml,which was significantly less than the 3D PC-ASL patients with perfusion defects,and the median volume of insufficient perfusion on the DSC was 180.2 ± 89 ml.Conclusion:3D PC-ASL can better reflect the larger perfusion defect and perfusion/diffusion mismatch area,which is consistent with DSC.Our results suggest that 3D PC-ASL scans are sufficient for screening patients with acute stroke who are contraindicated with contrast media.