Comparison of wavelet based coders applied to 3D Brain Tumor MRI Images

In this work, we compare the performances of three wavelet based coding algorithms such as 3D QTL, 3D SPIHT and JPEG2K applied to 3D brain tumor MRI images. An objective comparison is achieved between the three coders, namely 3D SPIHT, 3D QTL and JPEG2K. For those tests, eight MRI brain tumor scan test sets of 256 times 256 times 124 voxels of different patients have been used in order to compare those three coders. Results show the superiority of 3D SPIHT algorithm, whereas 3D QTL outperforms JPEG2K. Thus we may conclude, although 3D SPIHT perform good results, but it is not sufficient to reduce the storage space and transmission bitrate in case of telemedicine. So we must find a solution which can compress and transmit only the region important for diagnosis (in our case a tumor) with a high bitrate and the other regions with a low bitrate. The region of interest (ROI) based compression may be the solution to this problem.