EOS^(TM)三维影像评价成人脊柱侧凸的可靠性和准确性研究

Accuracy and reliability study of EOS^(TM) in the evaluation of adult scoliosis

目的研究EOS影像系统在成人脊柱侧凸评价中的可靠性和准确性。方法选择2015年9月至2016年1月,于我院进行EOS影像摄片的脊柱侧凸患者共41例。其中男5例,女36例,平均年龄64.6岁。利用ster EOS软件三维模型重建获得脊柱侧凸的各个测量数值,包括:Cobb’s角、顶点椎体轴面旋转角、T_(1~12)后凸角、T_(4~12)后凸角、L_(1~5)前凸角、L_1~S_1前凸角、骨盆入射角和矢状面骨盆倾斜角、外侧骨盆移位以及骨盆轴面旋转度。由2名放射科医师分别独立重建,并共同对模型与影像契合度进行评分。利用观察者间变异系数评价2名医师测得数据的观察者间变异。同时根据Cobb’s角将患者分为3组,并分析各组间契合度评分有无差异。结果脊柱测量值中,Cobb’s角、顶点椎体轴面旋转角、T_(4~12)后凸角、L_(1~5)前凸角和L_1~S_1前凸角的观察者间组内相关系数(intraclass correlation coefficient,ICC)分别为0.990、0.911、0.971、0.972和0.949(P<0.01)。仅T_(1~12)后凸角的观察者间ICC较差,为0.194(P=0.104)。骨盆测量值中,骨盆入射角、骶骨倾斜角、矢状面骨盆倾斜角、外侧骨盆移位以及骨盆轴面旋转度的观察者间ICC分别为0.846、0.885、0.914、0.987和0.779(P<0.01)。3组Cobb’s角不同的患者模型契合度评分差异有统计学意义(F=4.721,P=0.015)。结论使用EOS系统摄片和sterEOS后处理软件对成人脊柱侧凸进行3D建模能获得很好的测量可靠性和准确性,为医师提供了一种全新的方法以更好地评判脊柱侧凸。

Objective To study the accuracy and reliability of EOS^TM imaging system in the evaluation of adult scoliosis. Methods Between September 2015 and January 2016, EOS images from 41 adult scolisis patients（ 5 males, 36 females, mean age 64.6-year-old） were collected for the study. Ster EOS software was used to do 3D reconstruction of the spine to obtain a set of measurements, including Cobb＇s angle, apical vertebral rotation, T（1-12） kyphosis, T（4-12） kyphosis, L（1-5） lordosis, L1- S1 lordosis, pelvic incidence, sacral slope, sagittal pelvic tilt, lateral pelvic tilt and pelvis axial rotation. Two radiologists were blinded to each other to conduct 3D software generated measurements from an EOS imaging system. They also scored the conformance of 3D model and images. Intraclass correlation coefficients were compared to evaluate interobserver variability. Patients were also classified to 3 groups according to Cobb＇s angle to analyze conformance. Results Intra-class correlation coefficients were 0.990, 0.911, 0.971, 0.972 and 0.949（ P 〈0.01） for Cobb＇s angle, apical vertebral rotation, T（4-12） kyphosis, L（1-5） lordosis and L1- S1 lordosis, respectively. Only poor intra-class correlation coefficient： 0.194（ P = 0.104） for T（1-12） kyphosis. Intraclass correlation coefficients were 0.846, 0.885, 0.914, 0.987 and 0.779（ P 〈0.01） for pelvic incidence, sacral slope, sagittal pelvic tilt, lateral pelvic tilt and pelvis axial rotation. Conformation scores were statistically different among patients in 3 groups of different Cobb＇s angles（ F = 4.721, P = 0.015）. Conclusions EOS imaging and ster EOS 3D reconstruction create accurate and reliable measurements in adult scoliosis, providing a novel method for radiologists and orthopedists to improve the assessment of scoliosis.