CUDA-Accelerated Volume Rendering of Two-phase Bone

In this paper we present a new method to volume render two-phase bone more accurately. Firstly, we transform Computed Tomography (CT) value to gray value, and by using the method of without re-initial level set method we get the CT value of two-phase bone interface. Then, we implement the ray casting algorithm in graphics processing unit (GPU) with Compute unified device architecture (CUDA). And base on the CT value of two-phase bone interface, two-phase bone can be obvious separated with different colors. Experiments demonstrate that this method is useful and efficient.