检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:王雷[1,2,3] 高欣[1] 崔学理 梁志远[4]
机构地区:[1]中国科学院苏州生物医学工程技术研究所,江苏苏州215163 [2]中国科学院长春光学精密机械与物理研究所,吉林长春130033 [3]中国科学院大学,北京100049 [4]首都医科大学生物医学工程学院,北京100069
出 处:《光学精密工程》2014年第10期2815-2824,共10页Optics and Precision Engineering
基 金:国家自然科学基金资助项目(No.81371640);苏州市国际合作项目(No.SH201210);江苏省临床医药专项资金资助项目(No.BL2012049)
摘 要:针对影像导航手术提出了一种基于灰度距离融合的2D/3D刚性图像配准方法。该方法使用一种新的灰度距离信息对最常使用的传统相似性测度(互信息,互相关及模式强度)进行约束,构建一类新的相似性测度(距离互信息,距离互相关和距离模式强度),并用维也纳医科大学公开发表的2D/3D刚性配准金标准数据来评估新测度的配准性能。与基于灰度的传统相似性测度相比,文中方法构造的新相似性测度在平均目标配准误差(mTRE)的均值和标准差上均有显著降低,其中均值至少降低了28.15%,标准差最少降低了61.17%。以mTRE小于2mm为配准成功的依据时,新测度的配准成功率比传统测度至少提高了25.56%。此外,新测度在配准优化过程中的迭代次数比传统测度平均降低了35.59%。结果显示:基于灰度距离融合的2D/3D刚性配准方法比基于单一灰度的配准方法具有更好的图像配准性能。For image-guided surgery, a novel 2D/3D image rigid registration method is proposed by in- tegrating intensity distances of images. The method uses a new intensity distance information to re- strict the most commonly used similarity measures(Mutual Information (MI), Cross Correlation (CC) and Pattern Intensity (PI)) and to construct a kind of novel similarity measures(distance MI, distanceCC and distance PI). These novel measures are evaluated by using the porcine skull phantom datasets from the Medical University of Vienna. The experiments show that novel measures are better thantraditional measures, i. e. , the mean and standard deviation of mean Target Registration Errors (mTRE) by novel measures are respectively lower by at least 28.15% and 61.17% than those by tra-ditional measures. When setting mTRE less than 2 mm as successful registration, the success rate with novel measures increases by at least 25.56% on average. Meanwhile, the average iteration timesof novel measures also reduce by 35.59%than those of traditional measures. This results suggest that the novel registration method using novel measures has better performance of registration than intensi-ty-based methods using traditional measures in terms of the accuracy and robustness for 2D/3D rigid registration.
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.227