Purpose: To evaluate different rendering techniques of CT data for the assessment of long vessel segments in peripheral vascular occlusive disease.
Material and methods: 40 CT angiograms (aortoiliac: n = 20, leg arteries: n = 20) were viewed using three different rendering techniques: 1, maximum intensity projection (MIP); 2, volume rendering (VR); 3, shaded surface display (SSD). CT angiograms were obtained in 6 or 8 projections. Axial cross-section images were analysed using an interactive cine mode. Intraarterial DSA was the standard in all cases.
Results: The sensitivities for the diagnosis of occlusive disease were 100% (cross-section images), 94% (MIP), 91% (VR) and 93% (SSD). The specificities were 100%, 99%, 99% and 99%, respectively. For the accurate grading of high-grade (> 75%) stenoses, the sensitivities were 85% (cross-section images), 62% (MIP), 44% (VR) and 35% (SSD). Specificity was 99% for all techniques.
Conclusions: CTA is accurate in occlusive disease. Interactive viewing of cross-section images is the most accurate technique. MIP is superior to VR in the imaging of high-grade stenoses because contrast-to-noise ratio is high and thresholding is not necessary.