Embedded Wavelet Coding Of Medical Image With ROI

With the accelerating digitization of modern hospital, the digital medical images play a more and more important role in the diagnosis and treatment of human disease. Many hospitals have constructed the PACS system for archiving and communication of digital medical image. Because of the limited storage capacity and bandwidth, research on effective methods for compression and coding of the huge medical image data is critical for modern clinical practice.The coding of medical images is quite different compared with common images. Firstly, there are often important regions containing pathologic structures which are critical for diagnosis of disease. Also, there are background areas containing related issues that are less important. The general compression schemes are not suitable for medical image application. This paper proposes a novel medical image compression scheme. The key advantages include the following four points: the support of arbitrary shape ROI, smooth transition between ROI and background without border distortion, embedded lossy to lossless progressive refining code stream and random access to any ROI in the code stream. By using lossless compression on ROI with critical clinical details and high ratio lossy compression on background, the algorithm can achieve the goal of preserving the clinical relevant information of medical images as well as gains more compression ratio improvement than the traditional lossless scheme, and provides a good support for future object based applications.For the sequence images generated by CT and MR, this paper proposes a wavelet coding scheme based on three dimensional wavelet transform. The experiment results show some improvement over intra frame coding schemes. But improvements are still to be made on reducing the computation complexity and coding time to meet the requirements of clinical use.