目的 · 探讨CT、MRI 图像配准融合联合3D 打印技术，在辅助难治性骨盆肿瘤术前规划中的应用价值。方法 · 回顾性总结 2014 年 3 月至 2016 年 12 月于上海交通大学医学院附属第九人民医院接受治疗的难治性骨盆肿瘤病例，术前常规行薄层CT、MRI 检 查，借助Medraw 软件实现CT、MRI 图像快速配准融合，精确判定肿瘤安全边界，辅助设计并3D 打印个体骨盆肿瘤模型、假体和 （或）导板。结果 · 共纳入难治性骨盆肿瘤患者 13 例（软骨肉瘤 8 例、骨肉瘤合并动脉瘤样骨囊肿 1 例、未分化性肉瘤 1 例、纤维结 缔组织增生性纤维瘤 1 例、孤立性纤维瘤 / 血管外皮瘤 1 例、转移性鳞癌 1 例），平均年龄（50.0±8.5）岁。基于 Enneking 分区：Ⅰ区 2 例、Ⅰ + Ⅱ区 2 例、Ⅱ + Ⅲ区 1 例、Ⅰ + Ⅱ + Ⅲ区 3 例、Ⅰ + Ⅱ + Ⅳ区 1 例、Ⅰ + Ⅱ + Ⅲ + Ⅳ区 1 例、Ⅰ + Ⅳ区 1 例、Ⅱ + Ⅳ区 1 例、Ⅳ区 1 例。肿瘤最大直径平均值（15.15±4.81）cm，均未累及盆腔大血管。结合术中肿瘤边界快速冰冻结果及假体安装情况，判 断基于融合图像的术前评估、规划的虚拟手术方案与手术的实施一致。结论 · 薄层 CT、MRI 图像配准融合联合 3D 打印技术，可以精 确判定难治性骨盆肿瘤边界，辅助临床设计个体化 3D 打印假体和导板，在难治性骨盆肿瘤术前规划中具有重要的临床应用价值。

Objective · To explore the value of applying CT/MRI image registration and fusion combined with 3D printing technique in pre-surgical planning of refractory pelvic tumors.  Methods · A retrospective analysis was performed on the cases with refractory pelvic tumors admitted between March 2014 and December 2016 in Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine. CT/MRI examinations were performed as a routine pre-surgical procedure, and the resulting images were subject to rapid registration and fusion with the Medraw software in order to accurately determine the tumor boundaries and to be used as references in designing and making patient-specific tumor models, prostheses, and implants by 3D printing technique.  Results · A total of 13 cases with refractory pelvic tumors were included in this study (8 cases of chondrosarcoma, 1 case of osteosarcoma with aneurysmal bone cyst, 1 case of dedifferentiated sarcoma, 1 case of fibrous connective tissue hyperplasia, 1 case of solitary fibrous tumor/hemangiopericytoma, and 1 case of metastatic squamous cell carcinoma) with an average age of (50.0±8.5) years. According to the Enneking classification，there were 2 cases in zone Ⅰ , 2 in zones Ⅰ + Ⅱ , 1 in zones Ⅱ + Ⅲ , 3 in zones Ⅰ + Ⅱ + Ⅲ , 1 in zones Ⅰ + Ⅱ + Ⅳ , 1 in zones Ⅰ+Ⅱ+Ⅲ+Ⅳ, 1 in zones Ⅰ+Ⅳ, 1 in zones Ⅱ+Ⅳ, and 1 in zone Ⅳ . The mean maximal tumor diameter was (15.15±4.81) cm without implicating the pelvic blood vessels. Results from intraoperative frozen section at resection boundaries and the installation of prosthesis showed that the actual surgical procedure was consistent with the surgical planning based on the reported image fusion technique.  Conclusion · CT/MRI image registration and fusion technique combined with 3D printing technique can accurately determine the resection boundary of pelvic tumors and assist in designing of individualized prosthesis model and surgical guide, which has a great value in pre-surgical planning and clinical treatment of refractory pelvic tumors.