US Image-based Carotid Intima-Media Thickness Extraction And System Implementation

Atherosclerosis has gradually become one of the high risk diseases threatening human health.Because the cure rate of atherosclerosis is low and the recurrence rate is high,it is of great significance to detect and diagnose atherosclerosis in time to prevent the deterioration of the disease.Intima Media Thickness of carotid artery is an important indicator of atherosclerosis.Ultrasonic imaging technology is widely used in clinical practice.First,the ultrasonic image of carotid artery is obtained,but because of Speckle in the ultrasonic image,it is very difficult to extract the intima-media thickness of the carotid artery.And when the physician manually measures the IMT of the carotid artery,if the measurement data is large,it is not only time-consuming and laborious,but usually the result may be biased due to different measurement habits of the measurer.In order to solve the problem of impulse noise and unique speckle noise in ultrasound images,the median filtering and wavelet transform are studied separately,but they are not able to remove the noise in the ultrasound image well.To this end,this paper first improves the median filtering,and studies the improved extremum median filtering algorithm,and then combines with the wavelet transform to obtain a new filtering algorithm.The improved filtering algorithm of extreme value median filtering and wavelet transform proposed in this paper is used for image denoising.This combination of two image denoising algorithms for image denoising can preserve the edge information of the image well and has a good effect on removing noise in the ultrasound image.In order to accurately extract the IMT of the carotid artery,this paper separately studied the simulation data and clinical data,and introduced the carotid artery ultrasonic simulation based on Field II simulation platform.At the same time,this paper also studies the ultrasonic RF echo signals of clinical carotid arteries.The radio frequency image,envelope image and ultrasound image of the carotid artery are obtained by MATLAB ultrasonic simulation.Based on the study and analysis of the characteristics of carotid ultrasound images and the advantages and disadvantages of the existing methods,we find that the traditional Snake model can not extract the correct contours of narrow and deep concave regions.Image segmentation is not accurate enough and has certain limitations.To this end,this paper improves the traditional Snake model and studies the improved GVF-Snake model for image segmentation to extract the IMT of the carotid artery.Gradient Vector Flow external force field is used to improve the detection of middle film edge contour.First,the traditional Snake model is used to find the initial contour curve of the intima,and then the modified GVF-Snake model is used to converge the curve to the intimal edge,and the segmentation result of the intimal edge is translated downward by a certain distance as the initial contour of the middle membrane.The curve is converged to the mid-membrane edge using the modified GVF-Snake model.The experimental results show that the GVF-snake model proposed in this paper can accurately segment the carotid artery endothelium in ultrasound images.In addition,this paper uses the graphical user interface in MATLAB to design and develop the carotid artery ultrasonic simulation system.Users can freely select and process the RF signals of carotid artery ultrasonic echo they want to measure.This study can improve the diagnosis of carotid artery disease by medical staff,and provide convenience for the analysis of the disease.It is of great significance for the future study of carotid artery.