Research On Medical Image Compression Algorithm Based On PACS System

With the accelerated developing of hospital digital medical, the amount of medical imaging data grows dramatically, which affects the data storage space and access speed. In the field of signal processing, there is growing interest in the study based on the signal sparse expression. Signal sparse expression is not based on a fixed digital model of transformation, but the training set produced by the over-complete dictionary, by sparse coding the signal is broken down into a number of linear combinations of dictionary atoms, in order to obtain a more compact representation of the signal.Medical imaging data is a signal that this paper worked on picture archiving and communication system (PACS) in the study of the impact and Medical Digital Imaging and Communications (DICOM), based on the image data and applied research to obtain a preliminary exploration. For the sparse representation of the signal, from dictionaries to learn and study the expression of two aspects of the sparse research:learning dictionary main research direction algorithm are the method of optimal directions (MOD), K-singular value decomposition (K-SVD), iterative least squares based dictionary learning algorithms (ILS-DLA) and recursive least squares learning dictionary algorithm (RLS-DLA); sparse expression mainly studies basic pursuit (BP), matching pursuit (MP) and orthogonal Matching Pursuit (OMP). In the sparse representation signal after study, the use of mainstream dictionary learning algorithm acquired image data compression and decompression. In the compression and decompression process, analyze the impact of using different atomic scale influence dictionaries for generating a dictionary and its dictionary compression effect. Based on these studies, an improved compressed image K-SVD algorithm, which primarily used a dictionary of smaller scale atomic image compression and decompression, and saved the image edge information before decompression, after unpacking, the algorithm use the edge information image decompression after repair, the algorithm can achieve higher peak signal to noise ratio (PSNR) than the traditional compression algorithms.Finally, the proposed algorithm is added to a preliminary exploration of PACS. Demo program using the algorithm evaluated the time complexity of the algorithm is too large, after the need for subsequent improvement to the actual system.