CT肝脏灌注图像序列感兴趣区域运动跟踪算法研究  

RESEARCH ON ROI MOTION TRACKING ALGORITHM OF CT LIVER PERFUSION IMAGE SEQUENCE

作　　者：陈兆学[1] 赵晓静[1] 

机构地区：[1]上海理工大学医疗器械与食品学院,上海200093

出　　处：《计算机应用与软件》2013年第11期161-163,205,共4页Computer Applications and Software

基　　金：上海市教委科研创新项目(11YZ116);上海市教委PACS系统重点课程项目

摘　　要：为了有效修正CT肝脏灌注成像过程中各种运动干扰所造成的感兴趣区域位置的帧间偏差,提出一种基于mean shift方法和灰度直方图统计特征的肝脏图像序列感兴趣区域跟踪算法。基于该方法可以对CT肝脏灌注过程中肝门静脉为代表的感兴趣区域的中心位置的运动轨迹进行较为准确的帧间跟踪,从而有效修正CT灌注成像过程中感兴趣区域位置漂移所造成的TDC计算的误差。该方法可为临床上实施更为准确的肝功评价奠定基础。For effectively correcting the inter-frame offset of the ROI position caused by different motion disturbances during the course of CT liver perfusion imaging, an ROI tracking algorithm of CT liver image sequence based on mean shift method and grayscale histogram statistic characteristics is proposed. Based on this method it is able to carry out quite accurate inter-frame track on the motion trajectory in centre position of the liver portal vein （ a representative clinical liver ROI） during the course of CT liver perfusion, therefore effectively correct the error of TDC computation caused by the offset of ROI position during CT perfusion imaging process. This method can lay the foundation for performing more accurate evaluation of liver function in clinical applications.

关 键 词：CT肝脏灌注 时间密度曲线 均值漂移 感兴趣区域 运动跟踪 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]