摘要：

目的: 对比术前3D Slicer三维重建技术判断面肌痉挛责任血管与微血管减压术中实际所见的一致性。方法: 纳入2018年7月至2024年1月在山东省东阿县人民医院行微血管减压术的62例面肌痉挛患者，术前均行三维稳态进动快速成像（3D-FIESTA）和三维时间飞跃（3D-TOF）MRA检查，通过3D Slicer软件将3D-FIESTA和3D-TOF MRA图像配准并进行三维重建以明确责任血管，以术中实际所见为“金标准”，对比术前三维重建与术中实际所见的一致性。结果: 共62例患者中61例经术前三维重建明确责任血管，分别为小脑前下动脉47例、小脑后下动脉6例、椎动脉5例、椎动脉联合小脑前下动脉2例、小脑上动脉1例；59例与术中实际所见责任血管一致，1例术前三维重建考虑为椎动脉压迫，术中证实为椎动脉联合小脑前下动脉压迫，1例术前三维重建考虑为小脑前下动脉压迫，术中证实为小脑前下动脉联合迷路动脉压迫，1例未发现明显责任血管，术中证实为小动脉压迫。术前三维重建判断责任血管的准确率为95.16%（59/62），经一致性检验，与术中实际所见的一致性较高（κ=0.886，P=0.000）。结论: 面肌痉挛患者微血管减压术前通过三维重建技术明确责任血管具有很高的准确性，有助于制定手术方案，为术中责任血管的判断提供依据。

关键词: 痉挛, 面部肌肉, 微血管减压术, 磁共振成像

Abstract:

Objective: To compare the agreement in judging the responsible vessels for hemifacial spasm (HFS) between the preoperative 3D Slicer 3D reconstruction and the actual intraoperative judgment of microvascular decompression (MVD). Methods: A total of 62 patients with HFS who underwent MVD in Dong'e County People's Hospital of Shandong Province from July 2018 to January 2024 were included. 3D fast inflow with the steady state precession (3D-FIESTA) and 3D time-of- flight (3D-TOF) MRA were performed before surgery. 3D Slicer software was used to fuse 3D-FIESTA and 3D-TOF MRA images and 3D reconstruction to clarify the responsible vessels. The actual intraoperative judgment was taken as the "gold standard", and the agreement between preoperative 3D reconstruction and actual intraoperative judgment was compared. Results: There were 61 cases whose responsible vessels were identified by preoperative 3D reconstruction, including anterior inferior cerebellar artery (AICA) in 47 cases, posterior inferior cerebellar artery (PICA) in 6 cases, vertebral artery (VA) in 5 cases, VA + AICA in 2 cases, and superior cerebellar artery (SCA) in one case. The 59 cases were consistent with actual intraoperative judgment. In one case, the preoperative 3D reconstruction was considered as VA compression, and the intraoperative evidence was VA + AICA compression; in one case, preoperative 3D reconstruction was AICA compression, and the intraoperative evidence was AICA + labyrinthine artery compression; no significant responsible vessels were found in one case, and the intraoperative evidence was arteriole compression. The accuracy of preoperative 3D reconstruction was 95.16% (59/62). Agreement test showed a high consistency between preoperative 3D reconstruction and actual intraoperative judgment (κ = 0.886, P = 0.000). Conclusions: It is very accurate to define the responsible vessels by 3D reconstruction before MVD in patients with HFS, which is helpful to make the surgical plan and provide the basis for the judgment of the responsible vessels during the operation.

Key words: Spasm, Facial muscles, Microvascular decompression surgery, Magnetic resonance imaging