EOS影像三维重建在Chiari畸形伴脊柱侧凸患者中应用的可靠性及准确性

Reliability of EOS-3D reconstruction for scoliosis associated with Chiari malformation

目的:探讨使用EOS影像系统进行三维重建在Chiari畸形伴脊柱侧凸患者中应用的可靠性和准确性。方法:选取2017年7月~2018年4月于我院行EOS全脊柱正侧位片检查的Chiari畸形伴脊柱侧凸患者共14例,其中男5例,女9例,年龄9~18岁(13.4±2.9岁)。利用sterEOS软件分别对每例患者的EOS全脊柱正侧位片进行全脊柱三维建模并获得脊柱骨盆参数[包括脊柱侧凸Cobb角、顶椎旋转角(apical vertebral rotation,AVR)、T1~T12后凸角(T1~T12 TK)、T4~T12后凸角(T4~T12 TK)、L1~L5前凸角(L1~L5 LL)、L1~S1前凸角(L1~S1 LL)、骨盆倾斜角(PT)、骶骨倾斜角(SS)、骨盆入射角(PI)]的测量数值。由2名脊柱外科医师(医师A与B)分别独立重建,对各椎体进行契合度评分,并利用观察者间组内相关系数(intraclass correlation coefficient,ICC)来评价2名医师测得数据的观察者间变异。医师A同时在传统立位全脊柱正侧位X线片上测量上述参数,并将所测得参数与其第一次EOS三维重建所得参数进行一致性分析。3个月后2名医师分别对14例患者的EOS全脊柱正侧位片重新进行三维建模测量,对以上参数进行重测信度研究。根据脊柱侧凸Cobb角大小将患者分为3组:第1组Cobb角<40°,共3例;第2组Cobb角40°~60°,共7例;第3组Cobb角>60°,共4例。对契合度评分进行三组间两两比较,观察该系统在不同Cobb角大小患者中的应用情况。结果:EOS影像三维建模后,脊柱测量值中,Cobb角、AVR、T1~T12 TK、T4~T12 TK、L1~L5 LL和L1~S1 LL的观察者间ICC分别为0.980、0.990、0.987、0.945、0.990和0.963 (P<0.01);骨盆测量值中,PI、SS和PT的观察者间ICC分别为0.962、0.821和0.895 (P<0.01)。契合度评分中,第1组与第2组、第2组与第3组患者的评分之间无统计学差异(P>0.05),而第1组与第3组患者之间的评分存在显著性差异(P<0.05)。在传统立位全脊柱正侧位X线片与EOS三维重建全脊柱片的一致性分析中,脊柱侧凸Cobb角、T1~T12 TK、T4~T12 TK、L1~L5 LL、L1~S1 LL、PI、PT及SS的ICC值分别为0.981、0.991、0.843、0.975、0.972、0.997、0.998和0.996。2名医师的各参数重测可信度均很高,2名医师均在PT出现最小值,分别为0.883和0.807。结论:使用EOS影像系统摄片并进行三维重建对Chiari畸形伴脊柱侧凸患者进行影像学测量是可靠的,为脊柱侧凸的评估增添了一种新的方法。

Objectives:To investigate the reliability and accuracy of three-dimensional reconstruction by EOS imaging system in patients with Chiari deformity with scoliosis.Methods:Between July 2017 and April 2018,EOS images from 14 Chiari malformation associated with scoliosis patients(5 males,9 females,9-18 years old,mean age 13.4±2.9 years old) were collected.Ster-EOS software was used to perform 3 D reconstruction of the spine to obtain a set of measurements,including Cobb angle,apical vertebral rotation(AVR),T1-T12 kyphosis angle(Tl-T12 TK),T4-T12 kyphosis angle(T4-T12 TK),L1-L5 lordosis angle(L1-L5 LL),Ll-S1 lordosis angle(Ll-Sl LL),pelvic incidence(PI),sacral slope(SS),and pelvic tilt(PT).Two spinal surgeons independently reconstructed the vertebrae and evaluated the interobserver variability by using the interobserver coefficient of variation.Doctor A measured the above parameters obtained after first EOS-3 D modeling and the parameters in the traditional full spine image.And then consistency analysis of them was performed by calculating the intraclass correlation coefficient(ICC) value.Then two doctors studied the retest reliability of the above parameters.In addition,the patients were divided into three groups according to the size of scoliosis Cobb angle:the first group of Cobb angle <40°,a total of 3 cases;the second group of Cobb angle 400 to 60°,a total of 7 cases;the third group of Cobb angle>60°,a total of 4 cases.The three groups were compared in pairs to observe the application of the system in patients with different cob angle sizes.After that,the retest reliability of the above parameters was studied by two doctors.Results:The ICC values were 0.980,0.990,0.987,0.945,0.990 and 0.963(P<0.01) for Cobb angle,AVR,T1-T12 TK,T4-T12 TK,L1-L5 LL,and L1-S1 LL,respectively.The ICC values were 0.962,0.821 and 0.895(P<0.01) for PI,SS and PT.For fit score,there was no significant difference between group 1 and group 2,group 2 and group 3(P>0.05),while there was a significant difference between group 1 and group 3(P<0.05).In consistency analysis,the ICC values of Cobb angle,T1-T12 TK,T4-T12 TK,L1-L5 LL,L1-S1 LL,PI,PT and SS were as follows 0.981,0.991,0.843,0.975,0.972,0.997,0.998 and 0.996.The reliability of retesting the parameters by the two physicians was very high,and 2 doctors both showed a minimum value of 0.883 and 0.807 at the pelvic tilt,respectively.Conclusions:The 3 D modeling of Chiari deformity with scoliosis by using EOS system and sterEOS post-processing software can obtain good reliability and accuracy of measurement,which adds a new method for the evaluation of scoliosis.