基于ST图的微循环血细胞自动跟踪与测量技术  

Automatic tracking and measurement of blood cells motion in microcirculation based on ST image

作　　者：陈远[1] 赵志敏[2] 刘磊[1] 

机构地区：[1]南京航空航天大学自动化学院,南京211106 [2]南京航空航天大学理学院,南京211106

出　　处：《东南大学学报（自然科学版）》2011年第1期72-76,共5页Journal of Southeast University：Natural Science Edition

基　　金：国家自然科学基金资助项目(10172043);教育部博士点基金资助项目(20040287012)

摘　　要：为了有效分析微循环中血细胞的运动,利用血管中心线生成ST图,通过提取ST图中的轨迹,实现对血细胞的自动跟踪与测量.首先,设计出多尺度的方向滤波器,对ST图进行增强预处理;然后,在分析增强图像的概率密度分布函数和方向角度的基础上,设计了噪声滤波函数和方向滤波函数以提取细胞轨迹;最后,对提取的轨迹细化并计算其方向,实现对细胞的跟踪和流速测量.分别对人体微循环中的红细胞和白细胞进行跟踪与测量,跟踪的正确率达到96.5%以上,误差率小于1%.将流速测量结果与人工测量结果相比较,平均相关系数为0.98,高于现有的测量方法,表明该方法能更有效地分析和测量微循环中血细胞的运动.In order to analyze the blood cells motion in microcirculation effectively,a spatiotemporal（ST） image is generated according to the centerline of the blood vessel,and the blood cells tracking and measurement is realized by extracting the traces on the ST image.Firstly,a multi-scale directional filter is designed for ST image enhancement.Then,with the analysis of the probability density and the orientations of the enhanced image,the noise suppression function and the orientation filtering function are designed for extracting the traces.Finally,by thinning the extracted traces and calculating the orientations,the blood cells tracking and velocity measurement are realized.The red blood cells（RBCs） and leukocytes in microcirculation are taken as the experimental targets.The correct-tracking rate is more than 96.5% while false-tracking rate is less than 1%.Comparing the velocity measurement results with the manual evaluation,the average correlation coefficient is 0.98,which is higher than the existing method.The results show that the novel method can analyze and measure the blood cells motion in microcirculation more efficiently.

关 键 词：微循环 细胞跟踪 流速测量 时空图 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]