Study On Noise Analysis And Denoising Method Of Medical Ultrasound Images

Medical ultrasound images have been widely used for their advantages of low cost and no radiation.But limited by its imaging principle,the image resolution of medical ultrasound images is very low,and ultrasound images have serious noise pollution.Noise in medical ultrasound images is mainly the multiplicative speckle noise.It is an urgent problem to analyze the speckle noise distribution characteristics in medical ultrasonic images,and design some algorithm to eliminate it.This article use real medical ultrasound images to study the distribution characteristics of speckle noise,and then base on the neurobiological mechanism of visual simulation mechanism of biological vision adaptive denoising,designed the adaptive mechanism based on biological vision of medical ultrasonic images denoising method.So far,there have been many researchers starting from the principle of speckle noise in ultrasound images produced,through a series of conditions assume that the distribution of the speckle noise in medical ultrasound images to study its features,but lack researchs from the real medical ultrasound images to study the distribution of speckle noise.This article gets a lot of real three parts of the liver,breast and thyroid ultrasound images from the West China Hospital ultrasound,then cut many noise pixel blocks from these ultrasound images,take statistical tests for these noise pixels with K-S(the Kolmogorov Smirnov test)and the Goodness of Fit(the Goodness of Fit test)statistical tests.In this paper,it is found that the speckle noise in medical ultrasonic images is the noise that satisfies the gamma distribution.The presence of speckle noise reduces the clarity of medical ultrasound images,which in turn makes it more difficult for doctors to diagnose and cure diseases.The biological vision system can not only convert the received optical signals into nerve impulses,but also have a strong ability of information processing,which can be adapted to remove the noise in the image.Therefore,this article is based on the retina in the vertical information access mechanism,respectively to simulate a bipolar,adaptive mechanism of ganglion cells receptive field in adaptive to reduce the speckle noise in medical ultrasound images,at the same time keep the image details are clear,which can effectively improve the doctor's diagnosis accuracy.By analyzing the experimental results shows that this article's method that based on the mechanism of biological vision adaptive medical ultrasonic image denoising method not only can effectively restrain noise,can also enhanced image edge and other details.Compared with the classical anisotropic diffusion filtering,bilateral filtering,guided filtering and BM3 D denoising algorithm,it is proved that this method has better processing effect.