摘要
手工测量Cobb角是评价脊柱侧弯的金标准,但其测量中存在端椎和端线定位困难的问题.为此,提出一种基于掩膜分割的计算机辅助测量方法.随机选取134例脊柱侧弯患者X光数字图像,由2名经验丰富的人员采用传统手工方法测量其Cobb角,将其结果作为检验标准.利用所提改进的Cobb角测量方法对同一组图像进行测量.结果表明,改进的Cobb角测量方法与手工方法测量的Cobb角相比,具有良好的结果一致性.该方法可有效解决端椎和端线定位困难的问题,提升了测量可靠性和准确性.
Manual measurement of Cobb angle is the gold standard for the evaluation of scoliosis,however,there are some problems in locating end vertebrae and end lines.This paper proposed a computer-aided measurement,an improved Cobb angle measurement based on mask segmentation.The Cobb angles of 134 randomly selected digital X-ray images of patients with scoliosis were measured using traditional manual method,and the results were taken as the test standard.The improved measurement proposed by this study was used to measure the same images.Results show that the improved Cobb angle measurement method is consistent with that of the manual method.With the improved Cobb angle measurement,the problem of locating end vertebrae and end lines is solved,and the reliability and accuracy of measurement is improved.
作者
刘晓民
王哲
郭伟
吴承羽
张培水
刘仕佳
LIU Xiaomin;WANG Zhe;GUO Wei;WU Chengyu;ZHANG Peishui;LIU Shijia(College of Information and Communication Engineering,Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;Department of TCM Manipulative Orthopedic,Air Force Medical Center,Beijing 100142,China)
出处
《北京工业大学学报》
CAS
CSCD
北大核心
2021年第11期1284-1291,共8页
Journal of Beijing University of Technology
基金
总后重点项目(BKJ13j004)
空装重点项目(KJ20182A030135)
首发基金资助项目(2018-2-5121)。
关键词
脊柱侧弯
COBB角
脊柱分割
模式识别
边界提取
端椎定位
端线定位
scoliosis
Cobb angle
vertebra segmentation
pattern recognition
boundary extraction
end vertebrae location
end lines location
