Evaluating intracranial artery dissection by using three-dimensional simultaneous non-contrast angiography and intra-plaque hemorrhage high-resolution magnetic resonance imaging: a retrospective study.

Abstract

There was no previous report on the three-dimensional simultaneous non-contrast angiography and intra-plaque hemorrhage (3D-SNAP) magnetic resonance imaging (MRI) sequence to diagnose intracranial artery dissection (IAD). To improve the diagnostic accuracy and guide the clinical treatment for IAD by elucidating its pathological features using 3D-SNAP MRI. From January 2015 to September 2018, 113 patients with suspected IAD were analyzed. They were divided into IAD and non-IAD groups according to the spontaneous coronary artery dissection (SCAD) criteria. All patients underwent 3D-SNAP, 3D-TOF, T2W imaging, 3D-PD, 3D-T1W-VISTA, and 3D-T1WCE) using 3.0-T MRI; clinical data were collected. The IAD imaging findings (intramural hematoma, double lumen, intimal flap, aneurysmal dilatation, stenosis, or occlusion) in every sequence were analyzed. Receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic efficiency of each sequence. There was a significant difference in the probability of intramural hematoma, relative signal intensity of intramural hematoma, double lumen, stenosis, or occlusion signs on 3D-TOF, T2W, 3D-PD, 3D-T1W-VISTA, 3D-SNAP, and 3D-T1WCE sequences (P<0.05). The 3D-SNAP and 3D-T1WCE sequences were most sensitive for diagnosing intramural hematoma and displaying double-lumen signs, respectively. The diagnostic efficiency of the 3D-SNAP sequence combined with 3D-T1WCE was the highest (area under the curve [AUC] 0.966). The AUC value of the 3D-SNAP sequence (AUC 0.897) was slightly inferior to that of 3D-T1W enhancement (AUC 0.903). 3D-SNAP MRI is a non-invasive and effective method and had the greatest potential among those methods tested for improving the diagnostic accuracy for IAD.