【摘要】 针对互信息测度在配准医学图像时存在速度慢和有时误匹配的缺点,提出了2种改进型Iα信息测度:FNIα信息和SNLα信息。通过人体脑部CT/MR和MR/PET图像的刚体配准实验,从函数曲线、计算时间和收性能对V信息、互信息、Iα信息、FNIα和SNIα信息进行了比较与分析。实验结果表明,FNIα信息测度的收敛性能远优于其它测度,对噪声有很强的鲁棒性;SNIα信息在计算速度方面较互信息有一定程度的提高。更多还原

【Abstract】 In order to reduce the computational time and misregistration of mutual information in medical image registration,two improved Iα-information measures,named as FNIα information measure and SNIα information measure,were proposed.The function curves,computing time and convergence were studied by applying the following measures to rigid registration of computed tomorgrphy(CT)/magnetic resonance(MR) and MR/positron emission tomography(PET) images.The measures include V-information,mutual information,I-alpha information,FNIα information and SNIα information.The resutls of tests show that the FNIα information outperforms mutual information and other measures in convergence performance,and the computational of SNIα information is faster than that of mutual information.更多还原