中国组织工程研究 ›› 2024, Vol. 28 ›› Issue (9): 1372-1377.doi: 10.12307/2023.798

• 人工假体 artificial prosthesis • 上一篇    下一篇

全髋关节置换前人工智能(AI)规划的作用与意义

闵美鹏,吴  进,URBA   RAFI,张文杰,高  嘉,王云华,何  斌,范  磊   

  1. 南京医科大学第二附属医院骨科,江苏省南京市   210011
  • 收稿日期:2022-11-23 接受日期:2023-01-04 出版日期:2024-03-28 发布日期:2023-07-25
  • 通讯作者: 范磊,博士,副主任医师,南京医科大学第二附属医院骨科,江苏省南京市 210011
  • 作者简介:闵美鹏,男,1997年生,江苏省扬州市人,汉族,南京医科大学在读硕士,主要从事关节和创伤研究。 吴进,女,1978年生,江苏省镇江市人,汉族,2002年南京医科大学毕业,副主任护师,主要从事关节外科相关研究。
  • 基金资助:
    2022年江苏省卫健委老年健康科研项目(LKM2022008),项目负责人:范磊

Role and significance of artificial intelligence preoperative planning in total hip arthroplasty

Min Meipeng, Wu Jin, URBA RAFI, Zhang Wenjie, Gao Jia, Wang Yunhua, He Bin, Fan Lei   

  1. Department of Orthopedics, Second Affiliated Hospital of Nanjing Medical University, Nanjing 210011, Jiangsu Province, China
  • Received:2022-11-23 Accepted:2023-01-04 Online:2024-03-28 Published:2023-07-25
  • Contact: Fan Lei, MD, Associate chief physician, Department of Orthopedics, Second Affiliated Hospital of Nanjing Medical University, Nanjing 210011, Jiangsu Province, China
  • About author:Min Meipeng, Master candidate, Department of Orthopedics, Second Affiliated Hospital of Nanjing Medical University, Nanjing 210011, Jiangsu Province, China Wu Jin, Associate chief nurse, Department of Orthopedics, Second Affiliated Hospital of Nanjing Medical University, Nanjing 210011, Jiangsu Province, China
  • Supported by:
    Geriatric Health Research Project of Jiangsu Health Commission in 2022, No. LKM2022008 (to FL)

摘要:


文题释义:

人工智能(Artificial Intelligence,AI):是研究、开发用于模拟、延伸和扩展人的智能的理论、方法、技术及应用系统的一门新的技术科学。主要分支包括机器学习、神经网络、机器人技术、专家系统、模糊识别和自然语言处理,旨在利用计算机科学和强大数据模仿人类思维来制定决策和解决问题。
全髋关节置换术:将人工髋关节假体,包含髋臼侧和股骨侧,植入人体代替病变的髋关节,以矫正畸形、缓解关节疼痛、改善关节功能和提高患者生活质量,是治疗髋关节疾病最成熟可靠的手段。该手术适用于有严重髋关节疾患,伴有疼痛、关节活动受限,影响日常生活,且保守治疗无法缓解症状的患者。


背景:传统X射线胶片术前规划常常不准确,会导致一些术中和术后并发症,增加了手术时间和术中出血量,一定程度影响了全髋关节置换的手术效果。

目的:研究人工智能(AI)术前规划在全髋关节置换术中的准确性和应用效果。
方法:选取患侧初次行全髋关节置换的患者60例,其中30例应用人工智能(AI)三维术前规划为试验组、30例应用传统X射线胶片二维术前规划为对照组,两组患者的性别、年龄、病情等一般资料比较,差异均无统计学意义(P > 0.05)。比较两组患者术中实际安放假体与术前规划假体匹配情况、术中手术时间、术中出血量、术后即刻双侧股骨偏心距差值、双侧联合偏心距差值及双下肢长度差值和术后3个月Harris评分,分析两种术前规划的准确性和应用效果。

结果与结论:①两组患者均获得术后4-6个月的随访,其中对照组有1例患者于术后5 d发生假体后脱位,行手法复位后恢复,无再次脱出,其余患者均未出现术后并发症及术后死亡;②髋臼侧、股骨侧假体完全匹配率:试验组明显优于对照组(P < 0.05);③手术时间、术中出血量:试验组明显少于对照组(P < 0.05);④两组患者术后双下肢长度差值比较差异有显著性意义(P < 0.05),双侧股骨偏心距差值和双侧联合偏心距差值比较差异均无显著性意义(P > 0.05);⑤试验组患者术后3个月Harris评分明显高于对照组患者(P < 0.05);⑥结果说明,人工智能(AI)术前规划相较于传统胶片规划,更能准确预测假体型号、缩短手术时间、减少术中出血量、减少术后双下肢不等长发生、加快术后康复。

https://orcid.org/0000-0002-4548-0234 (闵美鹏);https://orcid.org/0000-0002-0236-9456 (吴进) 

中国组织工程研究杂志出版内容重点:人工关节;骨植入物;脊柱;骨折;内固定;数字化骨科;组织工程

关键词: 人工智能(AI), 全髋关节置换术, 术前规划, 模板测量, 假体

Abstract: BACKGROUND: The preoperative planning of traditional X-ray films is often inaccurate, which can lead to some intraoperative and postoperative complications, increase the operation time and intraoperative blood loss, and to some extent affect the surgical outcome of total hip arthroplasty.  
OBJECTIVE: To investigate the accuracy and effectiveness of artificial intelligence preoperative planning in total hip arthroplasty.
METHODS: Sixty patients who underwent primary total hip arthroplasty on the affected side were selected. 30 of them used artificial intelligence 3D preoperative planning (trial group) and 30 used conventional X-ray film 2D preoperative planning (control group), and there were no statistically significant differences between the two groups in terms of gender, age, condition and other general data (P > 0.05). The actual intraoperative prosthesis placement and preoperative planning prosthesis matching, intraoperative operation time, intraoperative blood loss, bilateral femoral eccentric distance difference, bilateral joint eccentric distance difference and bilateral lower limb length difference, and Harris score at 3 months after operation were compared between the two groups, and the accuracy and application effect of the two preoperative plans were analyzed. 
RESULTS AND CONCLUSION: (1) Patients in both groups were followed up for 4-6 months postoperatively. One patient in the control group had a posterior dislocation of the prosthesis at 5 days postoperatively, which recovered after performing manual repositioning without re-dislodgement. The rest of the patients did not have postoperative complications or postoperative death. (2) Complete matching rate of the prosthesis on the acetabular side and femoral side was significantly better in the trial group than that in the control group (P < 0.05). (3) Operation time and intraoperative blood loss were significantly less in the trial group than those in the control group (P < 0.05). (4) The difference in bilateral lower limb length between the two groups was statistically significant (P < 0.05), and the difference in bilateral femoral eccentric distance and bilateral joint eccentric distance was not statistically significant (P > 0.05). (5) Harris score of patients in the trial group was significantly higher than that in the control group 3 months after operation (P < 0.05). (6) These results confirm that compared with traditional film planning, artificial intelligence preoperative planning can predict the prosthesis type more accurately, shorten the operation time, reduce intraoperative blood loss, diminish the occurrence of postoperative bilateral lower limb inequality, and accelerate postoperative recovery.

Key words: artificial intelligence, total hip arthroplasty, preoperative planning, template measurement, prosthesis

中图分类号: