Cardiac MR imaging compression: comparison between wavelet based and JPEG methods

High quality compression on cardiac MR images using wavelet based methods have been compared with standard JPEG. Since the latter is an unrestricted algorithm designed for true-color realistic images, its performance compressing gray-level images can be improved with alternate algorithms specifically designed for this purpose. Here the authors present a method based on wavelet transforms and Huffman encoding; Daubechies and Lemarie-Battle wavelets have been used. Different combinations of parameters and transforms have been compared against JPEG with 100, 85, and 75 quality factor. The results show that wavelet based compression performed better in quality and compression rate; differences among the several wavelet based methods implemented have also been found.