人工智能三维术前规划在全髋关节置换术中的应用和准确性分析  

Application and accuracy analysis of artificial intelligence three-dimensional preoperative planning in total hip replacement

作　　者：丁冉[1] 王淇 刘烨 张启栋[1] 张念非[1] 郭万首[1] 王卫国[1] 张逸凌 Ding Ran;Wang Qi;Liu Ye;Zhang Qidong;Zhang Nianfei;Guo Wanshou;Wang Weiguo;Zhang Yiling(Department of Orthopedic Surgery,China-Japan Friendship Hospital,Beijing,100029;Peking University China Japan Friendship School of Clinical Medicine,Beijing,100029;Department of Surgical Anesthesiology,China-Japan Friendship Hospital,Beijing,100029;Longwood Valley Medical Technology Co.,Ltd.,Beijing,100190,China)

机构地区：[1]中日友好医院骨科,北京100029 [2]北京大学中日友好临床医学院,北京100029 [3]中日友好医院手术麻醉科,北京100029 [4]北京长木谷医疗科技有限公司,北京100190

出　　处：《生物骨科材料与临床研究》2022年第2期33-38,共6页Orthopaedic Biomechanics Materials and Clinical Study

基　　金：北京市自然科学基金资助项目(7202183);北京市科技计划课题(Z171100001017209)。

摘　　要：目的 探讨人工智能三维规划系统在全髋关节置换术中假体型号选择的准确性。方法 选取2019年11月至2020年12月住院期间的27例(32髋)行初次全髋关节置换手术患者,其中男10例,女17例。术前完成患侧髋关节X线及CT影像资料采集,分别使用传统模板测量法和人工智能三维术前规划进行假体型号预测,通过与全髋关节置换术中实际所用型号进行对比,比较两种预测方法的吻合率。结果 人工智能三维术前规划对髋臼及股骨假体型号预测的吻合率分别为90.6%和81.3%,传统模板测量对髋臼及股骨假体型号预测的吻合率分别为56.3%和46.9%。两种预测方法吻合率比较,差异有统计学意义(P<0.05)。人工智能三维规划系统预测髋臼杯型号同实际应用型号之间存在相关性(r=0.915,P<0.001),人工智能三维规划系统预测股骨假体型号同实际应用型号之间存在相关性(r=0.941,P<0.001)。结论 人工智能三维术前规划系统较传统模板测量方法更能准确预测假体型号。Objective To explore the accuracy of implant size in total hip arthroplasty using artificial intelligence-assisted three-dimensional planning system. Methods A total of 27 patients(32 hips) undergoing primary total hip arthroplasty were selected. The X-ray and CT image data were collected from November 2019 to December 2020 in China-Japan Friendship Hospital before the operation. Preoperative templating was performed for each total hip arthroplasty using both the traditional templating and artificial intelligence three-dimensional preoperative planning software. We analyzed the accuracy of implant size prediction using two methods. The coincidence rate of the two methods was compared. Results The predicted coincidence rates of acetabular and femoral prosthesis models by artificial intelligence three-dimensional preoperative planning were 90.6% and 81.3% respectively, which were more accurate than those of the traditional templating(56.3% for acetabular and 46.9% for femoral components). And the difference was statistically significant(P<0.05). There was significant correlation between artificial intelligence 3D planning software predicted preoperative acetabular implant size and intraoperative acetabular implant size(r=0.915, P<0.001), and 3D planning software predicted preoperative femoral implant size was significantly correlated with the intraoperative application of femoral implant size(r=0.941, P<0.001). Conclusion The artificial intelligence 3D preoperative planning software is more accurate than the traditional templating in predicting the implant size.

关 键 词：术前规划 人工智能 全髋关节置换术 

分 类 号：R687.4[医药卫生—骨科学]