摘要： 目的 评估放射成像中软骨下骨小梁纹理对膝骨关节炎（OA）进展的预测效能。方法 将OA公用数据库中，基线双侧膝关节Kellgren-Lawrence（KL）分级均为0级、随访36个月后右膝关节KL≥2级的研究对象纳入放射学进展组，36个月后KL分级仍为0级的研究对象纳入对照组。应用局部二值模式结合灰度共生矩阵的纹理分析方法，提取胫骨内侧平台软骨下骨小梁在不同角度和步长的纹理参数（对比度、相关性、能量、熵、同质性、异质性）。对有统计学差异的参数进行logistic回归分析，筛选能够预测OA发生的纹理参数，并对筛选后的纹理参数进行受试者操作特征（ROC）曲线分析，评估纹理参数的预测效能。结果 角度为0°~135°、步长为1~5的部分纹理参数与膝OA的发生存在相关性。ROC曲线显示联合协变量的各纹理参数的曲线下面积（AUC）为0.69~0.79，诊断比值比为5.75~15.6，当角度为45°、步长为1时异质性的AUC值最大，为0.79，诊断比值比为13.89。结论 基线膝关节X线尚无明显变化时，膝关节软骨下骨小梁的放射成像骨纹理参数能够预测OA的进展。

骨小梁纹理参数对膝骨关节炎进展的预测效能

孙英伟^1,2, 刘学勇³, 邓纯博⁴, 张展⁵, 马驯³

1. 辽宁中医药大学附属医院医学影像中心, 沈阳 110032;
2. 中国医科大学附属盛京医院放射科, 沈阳 110004;
3. 中国医科大学附属盛京医院康复科, 沈阳 110134;
4. 沈阳医学院附属中心医院运动创伤科, 沈阳 110024;
5. 中国医科大学附属盛京医院脊柱外科, 沈阳 110004

收稿日期:2021-05-16出版日期:2021-08-30发布日期:2021-07-29
通讯作者:刘学勇E-mail:liusjh@sj-hospital.com
作者简介:孙英伟(1982-),女,副主任医师,博士.
基金资助:辽宁省自然科学基金（20180530044）


关键词: 骨关节炎, 软骨下骨小梁, 局部二值模式, 灰度共生矩阵, 纹理分析
Abstract: Objective To evaluate the ability of radiographic subchondral trabecular bone texture in predicting knee osteoarthritis progression. Methods In the osteoarthritis database,the subjects who met the inclusion criteria were selected. The texture analysis method of local binary pattern combined with gray level co-occurrence matrix was used to extract the texture feature parameters under the medial subchondral cartilage of the tibial plateau,at different angles and steps. Texture parameters that predicted osteoarthritis occurrence were determined by performing logistics regression analysis. An area under the curve (AUC) analysis based on the receiver operating characteristic (ROC) curve was performed to evaluate the discrimination of texture parameters. Results Certain texture parameters with an angle in the range 0°-135° and step size in the range 1-5 were related to knee osteoarthritis occurrence. The AUCs of the texture parameters were in the range 0.69-0.79,and the diagnostic odds ratios were in the range 5.75-15.6. When the angle was 45° and the step size was 1,the maximum AUC value was 0.79 and the diagnostic odds ratio was 13.89. Conclusion When there is no significant change in the knee joint X-ray results at baseline,the radiographic bone texture parameters of the knee subchondral trabecular bone can predict the progression of osteoarthritis.
Key words: osteoarthritis, subchondral trabecular bone, local binary pattern, gray level co-occurrence matrix, texture analysis
PDF全文下载地址:

https://journal.cmu.edu.cn/CN/article/downloadArticleFile.do?attachType=PDF&id=2815