3D digital scan conversion of ultrasound image in volume rendering system

Abstract

Our volume rendering system consists of 3D digital scan conversion (DSC) part and volume rendering (VR) part, which requires large memory space to store 3D data and computation power to process them. Though there are several methods to reduce the computation amount such as early ray termination in VR, it is very difficult to process 3D DSC in real time. The 3D DSC converts the 3D sampled data in spherical coordinates or cylindrical coordinates to the 3D data in Cartesian coordinates. Because the input data of volume rendering system are obtained from 3D R-θ sampling, 3D R-θ interpolation has to be performed as the 3D DSC algorithm. However, 3D R-θ interpolation requires complicated operations, which requires large computation amount. This thesis proposes fast techniques of 3D DSC algorithms for 3D ultrasound images, which are fitted to volume rendering.