CT angiography: source images and postprocessing techniques in the detection of cerebral aneurysms.

Abstract

OBJECTIVE. The purpose of this study was to evaluate CT angiography source images and to assess various postprocessing techniques for cerebral aneurysm detection in daily practice. MATERIALS AND METHODS. We retrospectively studied 17 patients who underwent digital subtraction angiography and CT angiography for possible cerebral aneurysms. Four observers assessed CT angiography data for aneurysms, first after viewing source images in cine mode and again after consecutively adding postprocessing techniques in the following order: maximum intensity projection (MIP) of a 2-cm slab without bone editing (slab-MIP), taking 3 min of operator time; MIP with bone editing (edited-MIP), taking 20 min; shaded-surface display (SSD) without bone editing (3 min); and SSD with bone editing (20 min). In addition, the observers stated which postprocessing method they preferred. RESULTS. In nine patients, we found 13 aneurysms, 10 of which were larger than 5 mm in size. Sensitivity for detection of aneurysms on CT angiography was 92% for three observers and 77% for one observer. Specificity was 100% for two observers and 78% for two observers. Using four observers in a pooled analysis, the maximum of positive readings was 52. CT angiography was positive in 46 of these 52 instances: 41 positive readings after source image viewing, five additional positive readings after adding MIP, and no additional positive readings after adding SSD. Overall, MIP was judged superior to SSD and slab-MIP was judged to be better than edited-MIP. CONCLUSION. In this study interactive source image viewing detected most cerebral aneurysms. Additional slab-MIP is an efficient postprocessing technique for display much like that of angiography, for verification of aneurysms found, and for detection of additional aneurysms. These conclusions need to be verified in a larger study group.