Volumetric medical images visualization is an important tool in the diagnosis and treatment of diseases. Through history, one of the most dificult tasks for Medicine Specialists has been the accurate location of broken bones and of the damaged tissues during Chemotherapy treatment, among other applications; like techniques used in Neurological Studies. Thus these situations enhance the need of visualization in Medicine. New technologies, the improvement and development of new hardware as well as software and the updating of old ones for graphic applications have resulted in specialized systems for medical visualization. However the use of these techniques in mobile devices has been poor due to its low performance. In our work, we propose a client-server scheme, where the model is compressed in the server side and is reconstructed in a nal thin-client device. The technique restricts the natural density values to achieve good bone visualization in medical models, transforming the rest of the data to zero. Our proposal uses a tridimensional Haar Wavelet Function locally applied inside units blocks of 16x16x16, similar to the Wavelet Based 3D Compression Scheme for Interactive Visualization of Very Large Volume Data approach. We also implement a quantization algorithm which handles error coeficients according to the frequency distributions of these coe cients. Finally, we made an evaluation of the volume visualization; on current mobile devices .We present the speci cations for the implementation of our technique in the Nokia n900 Mobile Phone.