节点文献

基于计算机视觉的早期乳腺癌放射治疗一体化摆位系统研究

Study on integrated setup system with computer vision for early breast cancer radiotherapy

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 蔡新良郑业华龚剑朱超华贺志强

【Author】 CAI Xin-liang,ZHENG Ye-hua,GONG Jian,ZHU Chao-hua,HE Zhi-qiang(422nd Hospital of PLA,Zhanjiang Guangdong 524005,China)

【机构】 解放军第422医院器材科南方医科大学基础医学院南方医科大学生物医学工程学院

【摘要】 目的研究一种基于计算机立体视觉的乳腺癌放射治疗摆位验证的方法,以减少在乳腺癌分次放射治疗中,由于多次摆位、体型变化等因素的影响造成肿瘤位置动态位移引起治疗误差。方法使用由双摄像机组成的计算机视觉系统,准确匹配出多个特征标记物在左右两摄像机采集的图像中的具体坐标,依据双目成像的基本原理计算出各标记物在胸腹部表面特征点的三维坐标值,构建出乳房及胸部体表的立体特征形状,以此来完成在分次放射治疗前摆位准确性的在线判定。在算法设计过程中,采用动态选择待匹配图像和局部搜索的策略,目标图像匹配精确。结果调试实验表明,系统能够准确提供各标记点的三维空间位置,并且能够做到实时性计算。结论基于计算机立体视觉的乳腺癌放射治疗摆位系统能够为放射治疗摆位提供一定的定位数据指导。

【Abstract】 Objective To discuss a setup verification method based on computer stereo vision for breast cancer radiotherapy in order to reduce the treatment error caused by tumor dynamic displacement due to repeated setup and size change in breast cancer fractional irradiation.Methods The computer vision system was made up of two cameras.First,multiple characteristic markers’ coordinates were got by the two cameras,and then the three-dimensional coordinates of the markers were calculated according to the basic principle of binocular calculated.The stereo shape of breast and chest surface was constructed soon and the setup errors can be judge and corrected before radiation therapy.In addition,dynamic choice matching image and local searching strategy were used in the process of calculation and mergence in order to make the target image matching more accurate.Results Experimental results show that the system can accurately display the three-dimensional location of the markers,identify setup errors and can achieve real-time calculation.Conclusion The proposed method can reduce setup errors in breast cancer fractional irradiation and provide statistical support.

【基金】 国家重点基础研究发展(973)计划(2010CB732500);广东省科技计划项目(2010B031100013)
  • 【文献出处】 局解手术学杂志 ,Journal of Regional Anatomy and Operative Surgery , 编辑部邮箱 ,2012年01期
  • 【分类号】R737.9
  • 【被引频次】3
  • 【下载频次】62
节点文献中: 

本文链接的文献网络图示:

本文的引文网络