Research On Radial Artery Detection And Tracking Algorithm Based On Ultrasound Image

The radial artery is a terminal branch of the subclavian artery.By analyzing the radial artery,you can get a lot of important information about the human body,such as heart rate,blood flow rate,and blood vessel wall contraction.Medical staff at home and abroad generally use experience to identify the location of the radial artery for the diagnosis and treatment of the disease,which is time consuming and less accurate.In view of this situation,this paper innovatively proposes a method of automatically locating the radial artery in the ultrasound image using computer vision.The main work and innovations of the thesis are as follows:1.The denoising algorithm and morphological filtering algorithm of brachial artery ultrasound image have been studied.Firstly,the types of noise in the ultrasonic image are analyzed.Secondly,the basic principles of various denoising algorithms are studied,and the corresponding experiments are completed.The performance of different denoising algorithms in the brachial artery ultrasound image denoising scene was analyzed by the denoising algorithm evaluation index.The conclusion that the denoising performance of the threshold denoising algorithm based on wavelet transform is optimal in the brachial artery ultrasound image is obtained.At the same time,a variety of morphological operation principles were studied.Combined with the characteristics of brachial artery ultrasound images,the ultrasound images were closed.Experiments show that the use of denoising algorithm combined with closed operation can effectively remove noise,preserve wrist tissue details,Which lays the foundation for the subsequent radial artery detection algorithm.2.The principle of various edge detection operators was compared and analyzed.The performance of different operators in the edge detection of ultrasonic images was verified.The validity of the Canny operator edge detection algorithm in ultrasound images is proved by theoretical analysis and experiment;the principle of Freeman contour extraction algorithm is studied,and the extraction of the radial artery contour in the ultrasound image is realized.Based on the edge detection algorithm and contour extraction algorithm,a radial artery detection algorithm based on coordinate filtering and ellipse fitting algorithm and a radial artery detection algorithm based on frame difference method and ellipse fitting are proposed.The detection principles of the two algorithms are analyzed separately,and the performance and advantages and disadvantages of the two radial artery detection algorithms are compared and analyzed.The experimental results demonstrate the effectiveness of the two algorithms in different ultrasound image scenarios.3.Based on the advantages and disadvantages of two brachial artery detection algorithms,a radial artery detection and tracking algorithm combining two brachial artery detection algorithms and moving target tracking algorithm was proposed.The principle of Meanshift tracking algorithm is studied.The Meanshift-based radial artery tracking algorithm was designed and the corresponding experiment was completed.The experiment proves the limitation of the radial artery in the tracking ultrasound image.The principle of Kalman filtering algorithm is analyzed,and the radial artery tracking model based on Kalman filter and its algorithm principle are designed.The experimental results show that the algorithm can improve the stability of the brachial artery ultrasound detection algorithm,and the average detection rate of the radial artery is 94.5% when the ultrasound video is stable and the radial artery does not move violently.4.The brachial artery detection system based on ultrasound image was designed.The hardware platform of radial artery image acquisition and detection and tracking was established.The software framework of the brachial artery detection system based on ultrasound image was designed.Combined with the experimental results,the system's operation process and its algorithm processing flow are analyzed.The experiment shows that the system can realize the stable detection and tracking of the radial artery in the ultrasound image,which has innovation and engineering application prospects.