The Sequence Segmentation Algorithm For Coronary CTA Image

As the economy develops,people's lifestyles have changed dramatically.Cardiovascular and cerebrovascular disease poses a serious threat to human health due to high prevalence and mortality,with high mortality due to cardiovascular and cerebrovascular diseases each year.So the early diagnosis and prevention of cardiovascular is particularly important.As an effective noninvasive means for the diagnosis of coronary heart disease,multilayered spiral CTA has developed rapidly in recent years.VTA segmentation based on CTA images can accurately extract the coronary contours and is an important clinical assistive analysis tool for the diagnosis of coronary artery stenosis.The method can provide a quantitative analysis of the degree of calcification,plaque burden and stenosis.Therefore,it becomes is a Hot spots of medical image processing research field.And automatic or semi-automatic segmentation algorithm for coronary CTA sequence image has important clinical significance and practical value.The research work about the sequence of coronary artery segmentationis divided into three parts:Firstly,the sequence of the aorta is segmented.In this paper,a new algorithm based on ISODATA and region-growing sequence segmentation is used to segment the target of the aorta.The algorithm used ISODATA for clustering and the extracted center of the target area obtained by clusteringis set as the new seed point of the next CT image and then growing the region.The algorithm solves the problem of target splitting.Secondly,the sequence of coronary arteries is segmented.Coronary arteries in the CTA image target area are small,complex structure,so that automatic segmentation has some difficulty.Thus,this paper presents a tracking algorithm based on feature matching.Through the threshold of all the data,and then use the regional characteristics of the target match,and finally to achieve the sequence of coronary artery segmentation.The algorithm can well adapt to the identification and tracking of small area vessels.Thirdly,the improvement of coronary artery sequence segmentation algorithm.In this paper,an improved algorithm based on Kalman filter is proposed to overcome the shortcomings of missing coronary small blood vessels in the second part of the coronary sequence segmentation algorithm.The improvement consists of two parts: one is to replace the global threshold of the rough segmentation by using priori knowledge of the previous data under the guidance of a priori knowledge to calculate the local threshold.After that,taking into account the target blood vessel movement is too large lead to the ROI area is not the problem,the introduction of Kalman filter to predict the location as ROI regional center.The improved algorithm clearly identifies and tracks more coronary arteries and improves tracking accuracy.