期刊文献+

腰椎斜外侧椎间融合术后椎间角度变化的影响因素分析及预测模型构建

Factors affecting disc angle changes in oblique lateral interbody fusion:an analysis and predictive model development
原文传递
导出
摘要 目的探讨腰椎斜外侧椎间融合(oblique lateral interbody fusion,OLIF)术后椎间角度的变化及其影响因素,并建立椎间角度变化的预测模型。方法回顾性纳入2017年7月至2019年8月于北京积水潭医院脊柱外科因腰椎退行性疾病接受OLIF治疗的患者119例(共174个OLIF节段),男45例、女74例,年龄(62.1±9.8)岁(范围33~86岁)。OLIF术中使用的融合器前凸角度为6°。测量影像学参数:手术前后的椎间角度(disc angle,DA)及椎间角度变化值(ΔDA)、椎间高度(disc height,DH)、动力位椎间角度变化值(动力位ΔDA)、融合器位置和融合器倾斜度。通过Pearson相关分析确定ΔDA的影响因素,基于机器学习、采用十折交叉验证方法进行模型训练和验证,构建ΔDA的线性预测模型。结果174个节段术前DA为5.3°±5.0°,ΔDA为3.9°±4.8°;术前后缘椎体高度(PDH)为(6.6±1.9)mm,变化值(ΔPDH)为(3.1±2.1)mm;术后前缘椎体高度变化值(ΔADH)为(6.1±3.2)mm。Pearson相关性分析显示术前DA(r=-0.713,P<0.001)、动力位ΔDA(r=-0.153,P=0.044)和融合器位置(r=-0.183,P=0.016)与ΔDA呈负相关;ΔPDH与术前PDH显著负相关(r=-0.444,P<0.001),与融合器位置呈正相关(r=0.218,P=0.004)。术前DA<0°时ΔDA为10.8°±3.2°,术前DA为0°~6°时ΔDA为5.0°±3.7°,术前DA≥6°时ΔDA为1.0°±4.1°。通过十折交叉验证方法训练和验证模型,建立的预测模型公式为ΔDA=7.9°-0.8×术前DA,预测模型的R值为0.707,平均绝对误差为2.837。结论OLIF术后椎间角度的变化主要受术前椎间角度的影响,其次受融合器的前后位置的影响;通过机器学习建立的预测模型提示术前椎间角度可术前准确预测椎间角度的变化。 ObjectiveTo explore the factors affecting changes of disc angle(ΔDA)during oblique lateral interbody fusion(OLIF)and establish a predictive model ofΔDA.MethodsThis retrospective study included 119 patients with 174 segments undergoing OLIF procedures between July 2017 and August 2019 in Beijing Jishuitan Hospital.45 males and 74 females with an average age of 62.1±9.8 years(33-86 years)were included.The lordotic cages were all 6 degrees.Radiographic parameters included preoperative and postoperative disc angle(DA),disc height(DH),ΔDA on flexion-extension views(ΔDA-FE),cage location and cage inclination.Pearson correlation coefficient and machine-learning techniques were utilized to identify factors related toΔDA.Based on machine leaning techniques,ten-fold cross-validation for model training and validation were used to develop a predictive linear model forΔDA.ResultsThe averageΔDA was 3.9°±4.8°with preoperative disc angle(preoperative DA)of 5.3°±5.0°.The average change of posterior DH(ΔPDH)was 3.1±2.1 mm with preoperative posterior DH of 6.6±1.9 mm.The average change of anterior DH was 6.1±3.2 mm.Pearson correlation analysis showed a significant negative correlation betweenΔDA and preoperative DA(r=-0.713,P<0.001),cage location(r=-0.183,P=0.016),andΔDA-FE(r=-0.153,P=0.044).PDH changes were significantly negatively correlated with preoperative PDH(r=-0.444,P<0.001)and positively correlated with cage location(r=0.218,P=0.004).ΔDA was 10.8°±3.2°for negative preoperative DA(indicating kyphotic),5.0°±3.7°for preoperative DA between 0°and 6°,and 1.0°±4.1°for preoperative DA>6°.A predictive model was developed using ten-fold cross-validation,resulting in the formulaΔDA=7.9°-0.8×preoperative DA(R=0.707,MAE=2.837).ConclusionDisc angle changes in OLIF primarily depend on the preoperative disc angle,secondly on cage location.The predicting model based on machine-learning techniques using preoperative disc angle facilitates preoperative planning for OLIF procedures.
作者 吴静晔 葛腾辉 李观清 敖进涛 赵轩 孙宇庆 Wu Jingye;Ge Tenghui;Li Guanqing;Ao Jintao;Zhao Xuan;Sun Yuqing(Department of spine surgery,Beijing Jishuitan Hospital,Capital Medical University,Beijing 100035,China;School of Computer Science and Technology,Beijing Jiaotong University,Beijing 100044,China)
出处 《中华骨科杂志》 CAS CSCD 北大核心 2024年第18期1199-1206,共8页 Chinese Journal of Orthopaedics
基金 国家重点研发计划(2022YFC2407200,2022YFC2407204)。
关键词 脊柱融合术 放射摄影术 预测 机器学习 腰椎斜外侧椎间融合 Spinal fusion Radiography Forecasting Machine learning oblique lateral interbody fusion
  • 相关文献

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部