背景和目的：血管内再通是症状性颅内动脉粥样硬化性疾病（ICAD）的一种重要治疗手段。DSA是ICAD介入治疗后的随访手段，但其具有侵入性和辐射暴露等风险。因此在血管内治疗后的影像学随访中需要一种无创、准确的评估手段替代DSA。本研究旨在以DSA为参考评估高分辨率磁共振成像在随访中的图像质量和诊断效能。

材料和方法: 从2017年1月到2021年6月，在我们的前瞻性数据库中回顾性地收集了35名接受血管内再通治疗并接受HRMRI随访的40例颅内动脉粥样硬化性疾病患者的临床及影像学数据。对HRMRI的影像质量进行评估,并测量了DSA、TOF MRI及HRMRI影像上的参考血管及狭窄血管直径。使用Spearman相关系数，组内相关系数和Bland–Altman图对HRMRI的阅读者间一致性以及HRMRI与DSA的测量一致性进行分析。

结果: 本研究共纳入35名ICAD患者的40处病变。其中男性22例，女性13例，中位年龄63.6岁（IQR 58.5-70.0岁）。40处病变中8处病变接受支架置入手术，16处病变接受经皮腔内血管成形术（PTA），另16处病变接受药物涂层球囊（DCB）血管成形术。HRMRI在术后随访时成像质量优秀，整体图像质量为2.8±0.4级。在HRMRI图像上测量参考血管直径和狭窄血管直径时有着良好的阅读者间一致性，组间一致性评分分别为0.956和0.973。spearman相关系数表明无论是术前或者随访时，HRMRI与DSA都具有较好的一致性。HRMRI组参考血管和狭窄血管直径测量值与DSA的绝对（相对）差异分别为-0.040mm（1.4%）和-0.051mm（3.3%），优于TOF MRI组。此外，以DSA为参考标准，HRMRI对大于50%狭窄，大于70%狭窄和完全闭塞病变均有良好的诊断效能。

结论: HRMRI可以提供血管壁和管腔的高分辨率图像。基于HRMRI的管腔测量结果有着良好的阅读者间一致性。与TOF MRI图像相比，HRMRI图像测量管腔直径时与DSA绝对和相对差异更小。以DSA作为参考，HRMRI能够准确的评估随访时不同狭窄程度的病变。此外，HRMRI还可以显示管壁和斑块特征，为术后病情评估提供了更丰富的信息。

Background and Purpose: Endovascular treatment is an important treatment for symptomatic intracranial atherosclerotic disease (ICAD). DSA is a follow-up tool after ICAD intervention, but it carries risks such as invasiveness and radiation exposure. Therefore, a non-invasive and accurate assessment tools which is alternative to DSA is needed for imaging follow-up after endovascular treatment. The aim of this study was to evaluate the image quality and diagnostic efficacy of high-resolution magnetic imaging in follow-up using DSA as a reference.

Materials and Methods: From January 2017 to June 2021, data from 35 patients with 40 intracranial steno-occlusive lesions who underwent endovascular recanalization and received HR-MR follow-up were retrospectively collected in our prospective database.  The imaging quality of HRMRI was evaluated, and the reference vessel and stenotic vessel parameters on DSA, TOF MRI, and HRMRI images were measured. Inter-reader agreement of HRMRI and measurement agreement of HRMRI with DSA were analyzed using Spearman correlation coefficients, intra-group correlation coefficients and Bland-Altman plots.

Results: In total, 40 intracranial stenotic occlusive lesions in 35 patients with ICAD were included in this study. Of the 40 lesions, 22 were male and 13 were female, with a median age of 63.6 years (IQR 58.5-70.0 years). 8 of the 40 lesions underwent stenting, 16 underwent percutaneous transluminal angioplasty (PTA), and another 16 underwent drug-coated balloon (DCB) angioplasty. HRMRI had excellent imaging quality at postoperative follow-up, with an overall image quality of 2.8±0.4 grade. There was good inter-reader agreement between the reference vessel diameter and the stenotic vessel diameter measured on HRMRI images, with inter-group agreement scores of 0.956 and 0.973, respectively. spearman correlation coefficients indicated good agreement between HRMRI and DSA both preoperatively and at follow-up. The Bland-Altman plots are shown in Figure 4. The absolute (relative) difference between the reference vessel and stenosis vessel measurements and DSA and the absolute (relative) difference between the reference diameters in the HRMRI group were -0.040 mm (1.4%) and -0.051 mm (3.3%), respectively, which were better than those in the TOFMRI group. In addition, using DSA as the reference standard, HRMRI was good for greater than 50% stenosis, greater than 70% stenosis and completely occluded lesions.

Conclusion: HRMRI provides high-resolution images of the vessel wall and lumen. HRMRI-based vascular lumen measurements have good inter-reader agreement. Compared to TOF MRI images, HRMRI images show less absolute and relative differences from DSA when measuring lumen diameter. Using DSA as a reference, HRMRI can accurately assess lesions with different degrees of stenosis. In addition, HRMRI can also show the canal wall and plaque features, which provides richer information for postoperative disease assessment.