Li Jia, Mu Xiao-Dan, Zhang Yu-Jin, Zhao Bao-Gen, Wang Ning, Gao Ting, Zhang Li
Department of Interventional Medicine, Tangshan People's Hospital, Tangshan, 063001, China.
Department of Radiology, Peking University International Hospital, Beijing, 100080, China.
Eur Spine J. 2025 Mar 13. doi: 10.1007/s00586-025-08776-2.
To development a nomogram based on clinical features and apparent diffusion coefficient (ADC) of the cervical spinal cord in surgical prognosis in patients with cervical spondylotic myelopathy (CSM).
Patients with CSM who underwent decompression surgery between March and September 2023 were enrolled. Patients underwent conventional cervical spine MRI and sagittal position ZOOM-DWI before surgery. Recovery rate of neurological function was calculated based on the mJOA before and 6 months after surgery. According to recovery rate, patients were divided into good-recovery group (> 50%) and poor-recovery group (< 50%). Clinical- MRI factors model (Model 1) and Clinical-MRI-ADC factors model (Model2) were bulid by multivariate logistic regression to predict. Receiver operating characteristic curve (ROC), calibration curve, and decision curve analysis (DCA) were utilized to compare the predictive performance of the two models. A nomogram based on model 2 was constructed to predict poor recovery rate.
100 CSM patients were included in this study, including 60 patients in the good-recover group and 40 patients in the poor-recover group. Compared with model 1, the nomogram based on model 2 had a better AUC (0.933vs0.864). The calibration curve of model 2 is closer to the reference line, which indicates that model 2 has better resolution and accuracy. The DCA curve analysis of model 2 also showed better clinical utility. The nomogram based on model 2 performs well in predicting poor recovery rates.
The nomogram based on ADC values can effectively predict the outcome of postoperative neurological recovery in CSM patients.
基于临床特征和颈椎脊髓表观扩散系数（ADC）建立预测脊髓型颈椎病（CSM）患者手术预后的列线图。
纳入2023年3月至9月期间接受减压手术的CSM患者。患者术前接受常规颈椎MRI和矢状位ZOOM-DWI检查。根据术前和术后6个月的改良日本骨科学会（mJOA）评分计算神经功能恢复率。根据恢复率，将患者分为恢复良好组（>50%）和恢复不佳组（<50%）。通过多因素逻辑回归建立临床-MRI因素模型（模型1）和临床-MRI-ADC因素模型（模型2）进行预测。利用受试者操作特征曲线（ROC）、校准曲线和决策曲线分析（DCA）比较两个模型的预测性能。构建基于模型2的列线图以预测恢复不佳率。
本研究共纳入100例CSM患者，其中恢复良好组60例，恢复不佳组40例。与模型1相比，基于模型2的列线图具有更好的曲线下面积（AUC）（0.933对0.864）。模型2的校准曲线更接近参考线，表明模型2具有更好的分辨力和准确性。模型2的DCA曲线分析也显示出更好的临床实用性。基于模型2的列线图在预测恢复不佳率方面表现良好。
基于ADC值的列线图可有效预测CSM患者术后神经功能恢复情况。
