Guide Wire Of CT Image Detection Based On Tensor Voting

With the rapid increase of the number of patients with cardiovascular and cerebrovascular diseases at home and abroad,the treatment methods for cardiovascular and cerebrovascular diseases are also constantly optimized.Currently,vascular interventional surgery has become one of the important means to treat such diseases.In the process of this operation,the guide wire puncture intervention process is very important and indispensable.In the current clinical operation of vascular guidewire intervention,doctors mainly observe the local real-time CT image to carry out the whole process of guidewire intervention.In this kind of noisy CT image video to achieve a long time on the guide wire visual locking,for clinicians is a very easy process of visual fatigue.The content of this paper is to use the existing software technology to realize the guide wire assisted positioning function in the process of guide wire intervention.The research method is mainly to use tensor voting algorithm to extract the linear features of the guide wire in the gray scale CT image during the guide wire intervention process and carry out adaptive threshold analysis.In addition,DBSCAN clustering algorithm is combined to perform clustering analysis to eliminate other linear noises,so as to realize the extraction and positioning of the interventional guide wire in the CT image.The main research results are as follows:A.This is the first time to detect and locate the puncture guidewire in CT images during vascular interventional surgery.By the author fully research and analysis,the use of the needle thread in CT images in the strong linear significant this feature,coupled with tensor voting this strengthens the linear characteristics of an algorithm,combined with the design method of adaptive pixel value filtering and intensity of DBSCAN clustering algorithm after optimization of comprehensive treatment,made in this paper,the single frame thread detection of positioning has been very good.B.In view of the interference problem of linear noise in some CT images on linear feature extraction of guide wire,adaptive pixel value filtering analysis was carried out on the bar tensor feature map after tensor voting.Through the repeated adjustment of threshold feedback,only the sample points with high brightness such as guide wire in the range of 100-500 can be extracted from the linear feature graph after voting after filtering by pixel value.Then DBSCAN clustering analysis was performed to select the correct guide wire clustering point.