Automated bone editing algorithm for CT angiography: preliminary results.

Helical (spinal) CT has been shown to have a wide range of clinical advantages over dynamic CT in standard applications in the chest, liven, and pancreas [1-3]. These advantages have led to the exploration of its potential as a primary vascuIan imaging study. Recent reports have noted its possible role in vascular zones ranging from the circle of Willis and carotid arteries to the pulmonary arteries and thonacic aorta to the abdominal aorta, renal arteries, and portal venous system [411]. One of the prerequisite steps in CT angiognaphy, regardless of the three-dimensional imaging technique used, is removal of data not pertinent to the creation of the vascular image. In imaging of the abdomen this typically means removal of the nibs and spine. Although editing tools have been available for many years, they have often been cumbersome and inadequate when fine editing is needed. Quoted times for editing have varied but are typically noted to be anywhere from 1 5 mm to 2 hr. If CT angiognaphy is to be a routine examination, the editing task must be simplified and automated if possible. No general automated editing tool applicable in various parts of the body is available to our knowledge. Therefore, we have developed organ-specific segmentation methods for head and neck [1 2], chest [13], and abdomen [14]. Our abdomen segmentation technique is designed to automatically remove the spine and nibs so that the aorta, its major branching vessels, and the venous system can be viewed without obstructions in the edited data set. The purpose of this study was to determine the success of the editing tool. Materials and Methods