计算机辅助测量青少年特发性脊柱侧凸脊柱长度的方法及可行性分析

Accuracy and reliability of a computer-aided measurement method of spinal length in adolescent idiopathic scoliosis

中文关键词:  脊柱侧凸  青少年  脊柱长度  测量方法  计算机辅助

英文关键词:Scoliosis  Adolescent  Spinal length  Measurement method  Computer-aided

中文摘要:

  【摘要】 目的：提出一种新的计算机辅助测量青少年特发性脊柱侧凸（adolescent idiopathic scoliosis，AIS）脊柱长度的方法，评价其准确性和可重复性。方法：在前后站立位脊柱全长正位X线片上选取每个椎体的四个顶点，基于Matlab 6.5软件，设定程序确定L4下终板至T1上终板的中线，将中线上点的距离和作为脊柱长度。由1名医师采用计算机辅助测量方法和2名医师采用手工测量方法，随机选取5例AIS患者并对每例患者进行5次重复测量，计算两种方法测量结果的均值和标准差，评价计算机辅助测量脊柱长度方法的准确性和可重复性。选取45例AIS患者，其中女38例，男7例，年龄10～16岁（13.5±1.5岁）；Cobb角10°~45°（32.5°±8.7°）。由同一骨科医师在前后站立位脊柱全长正位X线片上分别采用手工测量方法和计算机辅助测量方法获得T1~L4脊柱长度，计算均值和标准差，比较两种测量方法的结果是否存在差异。应用计算机辅助测量方法测量计算出45例AIS患者L4~T1的脊柱长度，对同一医师应用该测量方法间隔4周的两次测量结果做比较，分析同一观察者在不同时间的可重复性；对3名不同医师分别应用该测量方法的测量结果做比较，分析该方法不同观察者间的可重复性。结果：应用计算机辅助测量方法测量5例AIS患者的脊柱长度（x±s）分别为325.8±0.46mm、391.3±0.43mm、476.7±0.04mm、441.3±0.79mm和432.8±0.50mm，2名医师应用手工测量方法测量5例AIS患者的脊柱长度分别为325.7±8.3mm、390.9±7.6mm、476.7±6.0mm、441.5±8.2mm、433.1±6.0mm和326.2±5.9mm、391.4±7.5mm、476.1±5.8mm、441.6±8.2mm、432.9±5.8mm，应用计算机辅助测量方法多次测量脊柱长度测得结果的标准差较手工测量结果低得多。采用计算机辅助测量方法与手工测量方法测量45例AIS患者的脊柱长度（x±s）分别为404.7±43.2mm和405.1±45.5mm，两组均数比较差异无显著性（P>0.05），采用计算机辅助测量方法测量AIS脊柱长度的同一观察者间隔4周的组内相关系数为0.995（P<0.05），不同观察者间的组内相关系数为0.942（P<0.05）。结论：计算机辅助测量AIS患者脊柱长度的方法较传统的手工测量方法操作简单，并且其准确性和可重复性均较好。

英文摘要:

  【Abstract】 Objectives: To evaluate the accuracy and reliability of a new computer-aided measurement method of spinal length in adolescent idiopathic scoliosis(AIS). Methods: Firstly, the method used to measure the spinal length was developed by Matlab 6.5, based on the digital photo of standing A-P radiographs of spine under the same condition. Using the program, four vertex points were selected from a vertebra, then a series of points were obtained on the center line. Based on these points, a spline interpolation algorithm was used to obtain smooth center line. Then, the spinal length could be calculated from these interpolated points. In order to assess the accuracy and reliability of the method, results of the means and standard deviations which measured five times repeatedly by one orthopedic surgeon using this method compared to results by two orthopedic surgeons using manual measurement method on five random digital photos. Forty-five patients(aged from ten to sixteen years old; 7 males, 38 females) with adolescent idiopathic scoliosis were selected, Cobb′s angle ranged from 10° to 45°, averaging at 32.5°. The spinal length of patients could be calculated by using this method. Three orthopedic surgeons performed the measurements independently and blindly in order to compare inter-observer reliability. And to compare intra-observer reliability, measurements were carried out two times on two different occasions by the same orthopedic surgeon. The interval between measurements was four weeks. Results: The measured results of five patients using computer-aided measurement method of spinal length were 325.8±0.46mm, 391.3±0.43mm, 476.7±0.04mm, 441.3±0.79mm, 432.8±0.50mm. And the measured results of the same five patients using manual measurement method by two different surgeons were 325.7±8.3mm, 390.9±7.6mm, 476.7±6.0mm, 441.5±8.2mm, 433.1±6.0mm and 326.2±5.9mm, 391.4±7.5mm, 476.1±5.8mm, 441.6±8.2mm, 432.9±5.8mm, respectively. Compared with the manual measurement results, the computer-aided measurement method of spinal length in adolescent idiopathic scoliosis was accurate, and the standard deviation of repeating measurement was lower. The spinal length of forty-five patients measured using computer-aided measurement method and mannual method was 404.7±43.2mm and 405.1±45.5mm respectively. There was no significant difference between the two methods through variance analysis or independent samples t test(P>0.05). Intraclass correlations for intraobserver at four weeks interval and interobserver agreements between three orthopedic surgeons in this measurement method of spinal length were 0.995 and 0.942 respectively(P<0.05). Conclusions: Compared to mannual method, the computer-aided measurement method of the spinal length in adolescent idiopathic scoliosis is accurate and reliable.