颈腰痛杂志2024,Vol.45Issue(4) :688-692.DOI:10.3969/j.issn.1005-7234.2024.04.020

基于腰椎矢状位及轴位MRI参数分析退行性腰椎滑脱的风险因素及构建预测模型

Analysis of the risk factors for degenerative lumbar spondylolisthesis based on lumbar sagittal and axial MRI parameters and construction of prediction model

陆祥平 熊壮 王保龙 周斌
颈腰痛杂志2024,Vol.45Issue(4) :688-692.DOI:10.3969/j.issn.1005-7234.2024.04.020

基于腰椎矢状位及轴位MRI参数分析退行性腰椎滑脱的风险因素及构建预测模型

Analysis of the risk factors for degenerative lumbar spondylolisthesis based on lumbar sagittal and axial MRI parameters and construction of prediction model

陆祥平 1熊壮 2王保龙 1周斌3
扫码查看

作者信息

  • 1. 枞阳县人民医院放射科,安徽铜陵 246700
  • 2. 安徽医科大学第一附属医院医学影像科,安徽合肥 230022
  • 3. 枞阳县人民医院骨二科,安徽铜陵 246700
  • 折叠

摘要

目的 基于腰椎矢状位及轴位MRI参数分析退行性腰椎滑脱(DLS)的风险因素并构建预测模型.方法 选取2021年1月至2023年9月本院收治的DLS患者70例(滑脱组),并按照滑脱组2倍的样本量选取同期无腰椎滑脱的腰椎退行性疾病患者140例(未滑脱组).两组患者均进行腰椎矢状位及轴位MRI扫描,比较两组MRI参数(腰椎间盘退变分级、小关节退变分级、终板Modic改变分型与腰骶角、黄韧带厚度、小关节积液宽度);采用有序多分类logistic回归模型分析DLS的影响因素,依据影响因素构建DLS发生的预测模型,并绘制受试者工作特征曲线(ROC)评价预测模型的预测效能.结果 与未滑脱组比较,滑脱组患者长时间负重占比高、腰椎间盘退变分级高、小关节退变分级高、终板Modic改变分型高、腰骶角大、小关节积液宽度大(P<0.05).Logistic回归分析显示,腰椎间盘退变分级>Ⅲ级(OR=13.730,95%CI:2.388~78.952)、小关节退变分级>Ⅰ 级(OR=11.025,95%CI:1.752~69.365)、终板 Modic 改变分型>Ⅰ 型(OR=7.386,95%CI:1.063~51.319)、腰骶角>45°(OR=17.127,95%CI:2.805~104.565)、小关节积液宽度>1.4 mm(OR=47.567,95%CI:2.858~791.612)是DLS发生的独立危险因素(P<0.05),无长时间负重是保护因素(OR=0.082,95%CI:0.009~0.758,P<0.05).根据上述因素构建预测模型,模型表达式为:Logit P=-8.339-2.507 ×无长时间负重+2.620 ×腰椎间盘退变分级+2.400 ×小关节退变分级+2.000 ×终板Modic改变分型+2.841 ×腰骶角+3.862 ×小关节积液宽度.Hosmer-Lemeshow检验显示模型拟合度较高(P=0.667),预测模型ROC曲线下面积为0.980,灵敏度为94.29%,特异度为95.79%(95%CI:0.961~0.999).结论 腰椎矢状位及轴位MRI参数中的腰椎间盘退变分级、小关节退变分级、终板Modic改变分型、腰骶角、小关节积液宽度及负重时间与DLS密切相关,根据影响因素构建的预测模型具有较好的预测效能,可为临床提供参考.

Abstract

Objective To analyze the risk factors for degenerative lumbar spondylolisthesis(DLS)based on lumbar sagittal and axial MRI parameters,and to construct a prediction model.Methods Seventy patients with DLS(DLS group)and 140 patients with-out degenerative diseases like lumbar spondylolisthesis(non-DLS group)who were admitted to the hospital from January 2021 to Sep-tember 2023 were selected.All patients received lumbar sagittal and axial MRI scanning.MRI parameters(lumbar disc degeneration grade,facet joint degeneration grade,endplate Modic modification of classification and lumbosacral Angle,thickness of ligamentum fla-vum,width of facet joint effusion)were compared between the two groups.Ordinal logistic regression analysis was conducted to screen the influencing factors of DLS,and a prediction model for DLS was constructed based on these factors.The predictive performance of the prediction model was evaluated with the receiver operating characteristic curve(ROC).Results Compared with non-DLS group,the proportion of long-term excessive weight bearing,grade of lumbar disc degeneration,grade of facet joint degeneration,class of Modic changes,umbosacral angle,and width of facet joint effusion in DLS group were higher/larger(P<0.05).Logistic regression analysis found that the grade of lumbar disc degeneration higher than grade Ⅲ(OR=13.730,95%CI:2.388-78.952),the grade of facet joint degeneration higher than grade I(OR=11.025,95%CI:1.752-69.365),Modic changes higher than class Ⅰ(OR=7.386,95%CI:1.063-51.319),lumbosacral angle larger than 45°(OR=17.127,95%CI:2.805-104.565)and width of facet joint effu-sion greater than 1.4 mm(OR=47.567,95%CI:2.858-791.612)were independent risk factors for DLS(P<0.05).No long-term excessive weight bearing was a protective factor(OR=0.082,95%CI:0.009-0.758,P<0.05).A prediction model was con-structed based on above-mentioned factors,and the expression was as follow:Logit P=-8.339-2.507 × no long-term excessive weight bearing+2.620 × grade of lumbar disc degeneration+2.400 × grade of facet joint degeneration+2.000 × Modic changes classi-fication+2.841 × lumbosacral angle+3.862 × width of facet joint effusion.Hosmer Lemeshow test showed a high degree of fitting(P=0.667).The area under the curve,sensitivity and specificity of the prediction model were 0.980,94.29%and 95.79%(95%CI:0.961-0.999).Conclusions The lumbar sagittal and axial MRI parameters(grade of lumbar disc degeneration,grade of facet joint degeneration,classification of Modic changes,lumbosacral angle,width of facet joint effusion and excessive weight bearing time)are closely related to DLS.The prediction model constructed based on the influencing factors has good predictive performance and can pro-vide reference for clinical practice.

关键词

退行性腰椎滑脱/磁共振成像/风险因素/预测模型

Key words

degenerative lumbar spondylolisthesis/magnetic resonance imaging/risk factor/prediction model

引用本文复制引用

出版年

2024
颈腰痛杂志
安徽医科大学

颈腰痛杂志

CSTPCD
影响因子:1.006
ISSN:1005-7234
段落导航相关论文