Integer Wavelet Transform For Medical Image Compression Research

With the quick development of digitalized medical pictures, PACS is more and more noticeable. The memory space of large capacity is needed to store the mass medical picture datum, and the size of the picture data is one of the factors, which decides the transmission speed while the picture is transmitted. Thus it is important to solve the archiving and transmission problem of medical pictures in PACS. The key is how to realize the effective compression of the medical pictures.The image compression standard in PACS now is the JPEG standard based on the DCT (Discrete Cosine Transform) algorithm, which can only be applied in the limited areas. With the birth and widely use of the JPEG2000, a new still image compression standard, we are interested in making full use of the wavelet transform in the medical image compression, which is the vital link in JPEG2000. The integer wavelet transform based on the Lifting Scheme for the medical picture compression is discussed in this paper. Firstly, the Lifting Scheme is adopted to realize the integer wavelet transform and reduce the operation complexity greatly. Secondly, on the basis of the research on the characteristics of the wavelet coefficients, the combination of the EZW (Embedded Zerotree wavelet) and advanced RLE (Run Length Encode) is presented to realize the efficient codes of the wavelet coefficients. Finally, the improved periodic extension algorithm is used to solve the boundary problem appeared in wavelet transform. The algorithm is realized by software and the good processing results are obtained.