基于MRI相关参数构建前交叉韧带胫骨止点撕脱骨折的列线图模型OA
Construction of a nomogram model of anterior cruciate ligament avulsion fracture at tibi-al insertion based on MRI-related parameters
目的 基于MRI相关参数构建前交叉韧带胫骨止点撕脱骨折的列线图(Nomogram)模型.方法 回顾性分析2019-01 至 2022-12 在武警江西总队医院骨科治疗的 143 例膝前疼痛患者的临床资料,根据影像学检查结果分为观察组(前交叉韧带胫骨止点撕脱骨折)和对照组(其他膝损伤患者或健康成人).使用多因素Logistic 回归分析前交叉韧带胫骨止点撕脱骨折发生的影响因素并构建患者前交叉韧带胫骨止点撕脱骨折风险的Nomogram预测模型,模型准确度以ROC曲线判断.另取 2023 年 1-8 月在医院治疗符合纳入标准的 70 例膝前疼痛患者进行成功率预测,与实际观测到的结果进行比较,并通过Hosmer-Lemeshow 检验评价该模型的拟合度.结果 143 例中,观察组 80 例(55.94%),对照组 63 例(44.06%),多因素分析结果显示,两组患者内侧胫骨平台后倾角[MTS,OR=1.733,95%CI(1.346～2.232)]、外侧胫骨平台后倾角[LTS,OR=1.907,95%CI(1.346～2.741)]及内侧胫骨平台深度[MTD,OR=0.059,95%CI(0.019～0.188)]为前交叉韧带胫骨止点撕脱骨折发生的独立影响因素(P<0.05).在此基础上建立风险预测模型.模型公式:Logit(p)=1.831+0.550×MTS+0.645×LTS-2.824×MTD.其ROC曲线下面积为 0.901,95%CI(0.849～0.952),敏感度为 82.50%,特异度为 88.90%,Youden指数为 0.714.外部验证模型预测总准确率85.20%,Hosemer-Lemeshow检验显示χ2=10.111,P=0.257.结论 基于MRI相关参数构建前交叉韧带胫骨止点撕脱骨折的Nomogram模型预测效能良好,MTS、LTS及MTD值能作为患者前交叉韧带胫骨止点撕脱骨折风险预测的指标.
Objective To construct a nomogram model of anterior cruciate ligament avulsion fracture at tibial insertion based on MRI-related parameters.Methods A retrospective study was conducted on 143 patients with anterior knee pain who were treated in the Department of Orthopaedics of Jiangxi Provincial Corps Hospital of Chinese People's Armed Police Force from January 2019 to De-cember 2022.According to the results of imaging examination,the patients were divided into an observation group(anterior cruciate ligament avulsion fracture at tibial insertion)and a control group(other knee injury patients or healthy adults).Multivariate logistic regression analysis was used to analyze the influencing factors of tibial avulsion fracture of cruciate ligament and to construct a nomo-gram prediction model for the risk of anterior cruciate ligament avulsion fracture at tibial insertion.The accuracy of the model was judged by ROC curve.Another 70 patients with anterior knee pain who met the inclusion criteria in the hospital from January to August 2023 were used for success rate prediction,the actual observed results were compared,and the fitting degree of the model was evalua-ted by Hosmer-Lemeshow test.Results Among the 143 cases,80 cases(55.94%)were in the observation group and 63 cases(44.06%)were in the control group.The Multivariate analysis showed that the medial tibial plateau posterior inclination angle[MTS,OR=1.733,95%CI(1.346,2.232)],lateral tibial plateau posterior inclination angle[LTS,OR=1.907,95%CI(1.346,-2.741)]and medial tibial plateau depth[MTD;OR=0.059,95%CI(0.019,0.188)]were independent influencing factors for the occurrence of anterior cruciate ligament avulsion fracture at tibial insertion(P<0.05).On this basis,a risk prediction model was es-tablished,whose model formula was Logit(p)=1.831+0.550×MTS+0.645×LTS-2.824×MTD.The area under the ROC curve was 0.901,95%CI(0.849,0.952),the sensitivity was 82.50%,the specificity was 88.90%,and the Youden index was 0.714.The total prediction accuracy of the external validation model was 85.20%,and the Hosemer-Lemeshow test showed χ2=10.111,P=0.257.Conclusions The Nomogram model of anterior cruciate ligament avulsion fracture at tibial insertion based on MRIp-related parameters has good predictive efficacy,and the MTS,LTS and MTD values can be used as indicators for predicting the risk of the disease.
付剑平;吴渊;万平印;陈树鑫;刘光强
330001 南昌,武警江西总队医院骨科330001 南昌,武警江西总队医院骨科330001 南昌,武警江西总队医院骨科330001 南昌,武警江西总队医院骨科330001 南昌,武警江西总队医院骨科
医药卫生
MRI前交叉韧带胫骨止点撕脱骨折列线图预测模型胫骨平台解剖参数膝关节生物力学
MRIanterior cruciate ligament avulsion fracture at tibial insertionnomogramprediction modeltibial plateau anatomical parametersknee joint biomechanics
《武警医学》 2026 (3)
235-240,6
评论