An Approach to Reduce Speckle Noise in Ultrasound Images

Abstract

Medical Image processing is one of the most fundamental tool since the inception of medical science. For detecting and curing any disease, image processing in medical played a very vital role. Ultrasound imaging brought about a huge stepping stone for medical testing and in many other field. But image disturbance has been the most backtracking effect in this case. Speckle Noising is one of the key disturbances and many have dedicated their efforts for denoising this defect Our main area of concern is to reduce this noise up to a requisite amount so that the image can be used without any problem. In our approach we targeted the properties of Speckle Noise and tried different approaches to reduce overall noises. We used Empirical Mode Decomposition(EMD) algorithm to our data (US Image) and observed the effect of EMD on such Data. In our work we used .rf DATA. We used the combination of SRAD, Wavelet Transformation (Double Density Discrete WT) and EMD. Then with the resultant data, we compared the image with different comparative parameters like MSE, PSNR and also observed their effect on lesion segmentation. For segmentation purpose we used watershed and seed growing method and obtained the result