Interactive Editing for 3D Ultrasound Volume Rendering

In 3D ultrasound volume rendering, the region of interest is blocked by near field acoustic data. Consequently, it is very hard to give a correct diagnosis from the 3D image. This paper presents two interactive editing methods to remove the unwanted data and improve the image's quality. The first method is in the 3D domain. It isolates ROI's interactively and removes the unwanted structures starting from an inverse mapping of the shear-warp procedure to identify samples to be removed in the volume data. Afterwards, one can see the updated 3D image without blocked structures. The second method utilizes adaptive border detection method starting from user defined ROIs on an arbitrary slice to a sequence of images propagated through the normal vector of the processing slices. The unwanted samples in the volume data can be removed slice-by-slice. Finally, a 3D image can be rendered with no blocked structure. We have built an interactive user interface to manipulate the proposed editing software as one of the 3D tools in our lab.