computer methods and programs in biomedicine 90 ( 2 0 0 8 ) 117–123 journal homepage: www.intl.elsevierhealth.com/journals/cmpb Real-time visualization of large volume datasets on standard PC hardware Kai Xie a,* , Jie Yang b , Y.M. Zhu c a The Methodist Hospital Research Institute, Texas Medical Center, 77030 Houston, TX, USA b Institute of Image Processing & Pattern Recognition, Shanghai Jiaotong University, 200240 Shanghai, China c CREATIS-CNRS Research Unit 5515 & INSERM Unit 630, 69621 Villeurbanne, France article info Article history: Received 21 March 2006 Received in revised form 7 October 2006 Accepted 12 December 2007 Keywords: Graphics processing unit Optic transfer function Early ray termination Empty-space skipping Visibility culling abstract In medical area, interactive three-dimensional volume visualization of large volume datasets is a challenging task. One of the major challenges in graphics processing unit (GPU)-based volume rendering algorithms is the limited size of texture memory imposed by current GPU architecture. We attempt to overcome this limitation by rendering only vis- ible parts of large CT datasets. In this paper, we present an efficient, high-quality volume rendering algorithm using GPUs for rendering large CT datasets at interactive frame rates on standard PC hardware. We subdivide the volume dataset into uniform sized blocks and take advantage of combinations of early ray termination, empty-space skipping and visi- bility culling to accelerate the whole rendering process and render visible parts of volume data. We have implemented our volume rendering algorithm for a large volume data of 512 × 304 × 1878 dimensions (visible female), and achieved real-time performance (i.e., 3–4 frames per second) on a Pentium 4 2.4 GHz PC equipped with NVIDIA Geforce 6600 graphics card (256 MB video memory). This method can be used as a 3D visualization tool of large CT datasets for doctors or radiologists. © 2007 Elsevier Ireland Ltd. All rights reserved. 1. Introduction In medical area, visualization of volumetric datasets acquired by computed tomography (CT), helps to understand the patient’s pathological conditions, improves surgical planning, and has an important role in education. However, a typical data size of today’s clinical routine is about 512 × 512 × 1024 (16 bit CT data) and will increase in the near future due to technological advances of acquisition devices. Conventional slicing is of limited use for such large datasets due to the enormous amount of slices. However, providing interactive three-dimensional volume visualization of such large datasets is a challenging task. * Corresponding author at: 1956 Dryden Road, Apartment 10, Houston, TX 77030-1392, United States. Tel.: +1 8327586051. E-mail address: xie kai2001@sina.com (K. Xie). In this paper, we present an efficient, high-quality volume rendering algorithm using graphics processing units (GPUs) for rendering large CT datasets at interactive frame rates on standard PC hardware. We employ the 3D texture mapping capability commonly available in modern GPUs as a core ren- dering engine and take advantage of combinations of early ray termination, empty-space skipping and occlusion clipping to accelerate the whole rendering process. Moreover, using the voxel value table and filtered visible blocks, we reduce the data traffic between CPU and GPU by eliminating the blocks that only have a small number of visible voxels to be rendered. As a result, we are able to achieve an interactive performance to render a large volume data of a 512 × 304 × 1878 dimension. 0169-2607/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.cmpb.2007.12.006