Carotid Artery Segmentation Algorithm Based On CTA Images

Stroke is one of the three leading causes of death or disability worldwide.About 2/3 of strokes are ischemic strokes.Carotid stenosis and occlusive disease are one of the most common causes of ischemic stroke.Early identification of the vascular morphology of the carotid artery is important for assessing the risk of stroke.Computed tomography enhancement(CTA)has been used in the diagnosis of carotid stenosis.The main difficulties in carotid artery segmentation based on CTA images are:(1)the number of slices is large;(2)there are more interfering factors(bones,similar vessels)around the carotid artery in CTA images;(3)the morphology of the common carotid artery is changing greatly at the bifurcatiori stage and the curvature of the carotid artery after bifurcation is large.These features will make the segmentation algorithm more difficult.Based on the above problems,this paper proposes two methods for rapid and stable segmentation of the carotid artery.Based on 8 groups of real clinical data from hospitals,the main research contents are as follows:(1)Regional growth method based on adaptive region features with multiple sub-points:Firstly,adaptive iterative CT values and morphological methods are used to remove soft tissue around the carotid artery and eliminate boundary adhesion;secondly,the carotid artery is unique when bifurcation The feature automatically determines the number and location of the seed points;finally,the carotid artery region is obtained by region growing according to the seed point and growth criteria.The experimental results show that the algorithm can accurately segment the internal carotid artery and the external carotid artery after carotid bifurcation.Compared with other regional growth algorithms,this method can better meet the clinical needs in terms of accuracy and time.(2)Segmentation method based on image enhancement combined with regional features:Firstly,the automatic update method of the region of interest(region containing carotid artery)is used to ensure that the region is as small as possible and the calculation amount is reduced.Secondly,the enhancement algorithm is used to highlight the blood vessel region.The eight-neighbor boundary tracking method is used to obtain the connected region boundary for the enhanced image.Finally,the connected region features are used to identify the carotid artery region.The experimental results show that the segmentation result obtained by the algorithm has higher accuracy and the algorithm running time is shorter.Compared with different enhancement algorithms and boundary extraction methods,the segmentation results obtained by this method are more accurate.This method not only avoids the shortcomings of the segmentation result based on the adaptive region feature multi-point region growth method,but also obtains the approximate centerline of the blood vessel.The innovation of this study lies in:(1)The selection of seed points in the regional growth method based on adaptive region features with multiple sub-points combines the characteristics of the carotid artery in the bifurcation stage to automatically calculate the number and position of the seed points,ensuring that the common carotid artery can still be segmented two carotid regions(intra-and extra-cervical arteries)after bifurcation;(2)Segmentation method based on combination of image enhancement and regional features can get the center of the area of interest in different stages of common carotid artery and intermal and external carotid arteries,which ensures that the region of interest always contains the carotid artery;(3)Eliminating other similar vascular regions by judging the distance between the center of gravity of the connected region and the central point of the region of interest is removing the interference factors around the carotid artery in the segmentation method based on image enhancement combined with regional features.