The Research Of Medical Image Compression Technique Based On Wavelet Transform

With the development of medical information technology, more and more medical imaging equipments are appearing in all levels of medical service organization,and a great number of medical image data have thus been produced,which poses higher requirement on medical image storage and transmission. Therefore, it is necessary to use valid image compression to lower the number of image data. Wavelet transform with characteristics of good localization in frequency time domains and multi resolution, have a wide application in medical image compression. The research subject of this paper is put_forward under this background, this paper contains the following parts:1 .The paper introduces the background and significance of the research subject and reviews researches on image compression home and abroad, it also discusses the basic theories and methods related to the object, including basic conception image compression,basic theories and methods related to the image compression and theory on wavelet transform.2.The paper studies the application of wavelet transform in medical image compression, analyze some key questions of medical image compression which is based on wavelet transform, including medical image waveform decomposition, choice properties of wavelet bases, and analysis of wavelet number and border expansion; Then analyze the Embedded Zerotree Wavelet method in detail, including it's thinking, principle of work and the processing of coding and decoding.3.Because of the decrease in the number of zero trees results from too big margin between quatification threshold values of original algorithm, a series of faults are caused.In light of this,a new quatification threshold value is adopted to improve the traditional EZW method,to relatively reduce the margin between quatification threshold interval and to increase the number of zero trees in the process of coding,and thus raises the efficiency of coding.4.Experiments were carried out with the use of simple wavelet coding method, original EZW method and improved EZW coding algorithm.The results of the experiments show that the MSE and PSNR of improved algorithm is better than that of original EZW at the same ratio.The last part of the paper concludes the paper,putting forward some research topics and possible solution methods for future study.