目的:探讨基于影像融合和三维表面重建影像后处理技术的术前模拟在颅内动静脉畸形(AVM)治疗中的应用价值。方法:回顾性分析2018年1月至2019年12月苏州大学附属第一医院神经外科手术治疗21例AVM患者的临床资料。术前均行头颅CT、3.0 T MRI和三维数字减影血管造影(3D-DSA)检查，利用iPlan软件将所有影像学数据配准融合并重建出三维模型，根据模型评估AVM的位置、范围、与功能区的关系以及供血动脉、引流静脉的位置，然后与术中所见进行比对，并在显微镜下切除AVM。术后行头颅MRI和DSA随访，同时检测患者的四肢活动、语言功能及视力情况。结果:21例患者均顺利完成术前模型的重建，其中20例利用模型在术中成功辨认AVM、功能区、供血动脉及引流静脉的位置；1例因术野内无明显粗大的血管且脑内血肿造成脑沟受压，脑沟回形态不清晰，无法完成比对。术前模型与术中评估的符合率为95.2%(20/21)。21例患者中，AVM全切除20例，次全切除1例。5例因出血导致肢体活动障碍的患者中，4例于术后3个月内恢复正常；1例因出血导致偏盲的患者，术后3个月复查视野缺损有所改善。1例因术后再出血，出现肢体活动障碍，于2周后恢复正常。21例患者的随访时间为(15.8±6.7)个月(3～25个月)，行影像学复查均未发现再出血。4例癫痫起病的患者中有3例术后无发作，1例于术后3个月内仍有发作，经调整抗癫痫药物治疗后发作停止。结论:利用影像后处理技术有助于在术前对AVM的位置、范围、与功能区的关系以及供血动脉、引流静脉的位置作出准确评估，可提高AVM的全切除率，减少术后并发症的发生。“,”Objective:To explore the application value of image post-processing technology based on image fusion and three-dimensional surface reconstruction in the treatment of intracranial arteriovenous malformations (AVM).Methods:A total of 21 patients with intracranial AVM underwent surgical treatment at Neurosurgery Department, the First Affiliated Hospital of Soochow University from January 2018 to December 2019 and their clinical data were retrospectively analyzed. All patients underwent the examinations of preoperative CT, 3.0T MRI and 3D-DSA. Multimodal images were then co-registered with software iplan, followed by three-dimensional image reconstruction to generate preoperative models. The location and range of AVM, relationship to eloquent area, and locations of feeding artery and draining vein were evaluated based on the model and compared with intra-operative findings. The AVM was resected under craniotomy microscope. Head MRI and DSA were followed up post operation, as well as the patient’s limb movement, language function and vision.Results:For all 21 patients, preoperative model reconstruction was successfully made. The location of AVM, eloquent area, feeding artery and draining vein were identified during operation using the model. Comparison could not be performed in one case, because there was no obvious large blood vessel in the operative field, and the hematoma in the brain caused compression on the brain sulcus and unclear shape of the gyri. The rate of consistency between preoperative modeling and intraoperative findings was 95.2% (20/21). Among the 21 patients, total resection was performed in 20 cases and subtotal resection in 1. Of the 5 patients with hemiplegia caused by AVM bleeding before operation, 4 recovered within 3 months. One patient with hemianopsia due to bleeding before operation reported improvement of the visual field at 3 months after operation. Postoperative dysfunction of limb movement was observed in 1 case because of rebleeding after the first surgery who and recovered to normal in 2 weeks. For 21 patients, the follow-up lasted for 15.8±6.7 months (3-25 months) and there was no case of rebleeding. Among the 4 patients with epileptic seizure as the initial symptom, 3 had no seizure after operation and 1 still had seizures within 3 months after surgery. The seizure stopped following the adjustment of antiepileptic drugs.Conclusion:The technology of image post-processing can help to make an accurate assessment of the position and range of AVM, its relationship to functional area, the location of feeding arteries and drainage veins before surgery, which can improve the total resection rate of AVM and reduce the occurrence of postoperative complications.