Carotid artery stenosis: optimization of CT angiography with a combination of shaded surface display and source images.

PURPOSE To evaluate the accuracy of CT angiography of occlusive disease of the carotid arteries using three-dimensional surface-rendered images alone and in conjunction with display of axial source images. METHODS Forty-eight symptomatic patients had conventional angiography followed by CT angiography within 24 hours. Images of 96 carotid arteries were acquired using contrast-enhanced spiral CT. Image postprocessing was performed on a free-standing workstation to produce 3-D shaded surface display (SSD) images. Two readers independently evaluated the CT angiographic (SSD) images and then reevaluated each case while simultaneously reviewing the SSD and axial source images. Digital subtraction angiograms were evaluated in a separate session and eventually compared with CT angiograms. All evaluations were performed under blinded conditions to control for reader bias. RESULTS SSD images alone underestimated stenosis relative to angiograms whereas combined SSD and axial images did not. CONCLUSION SSD angiograms in conjunction with the source images are better than SSD images alone in estimating the degree of stenosis in carotid artery disease.