Purpose: To analyze the detection of endoleaks with low-tube-voltage computed tomographic (CT) angiography.
Materials and methods: A cylindrical plastic phantom mimicking an aortic aneurysm, containing a stent graft and 36 simulated endoleaks of various diameters, densities, and distances from the graft, was placed into three different water containers corresponding to small, intermediate-sized, and large patients. CT was performed at 80, 100, and 120 kVp, and the volume CT dose index (CTDI(vol)) was noted. Three radiologists independently analyzed the CT images for endoleak localization with diagnostic confidence and for image quality. Analysis of variance was used to compare results and to find parameters that have an effect on endoleak detection.
Results: All 6-mm endoleaks were correctly identified at all tube energies except for five endoleaks at 80 kVp in the thick phantom. Observers detected more 4-mm leaks at higher tube voltage in the thick phantom (P < .001) and at 100 kVp compared with 80 kVp in the intermediate phantom (P = .006). Compared with CTDI(vol) at 120 kVp, CTDI(vol) was lower by nearly 50% at 80 kVp in the thin and intermediate phantoms and at 100 kVp in the thick phantom. Endoleak diameter, position, phantom dimension, and tube voltage all had a significant effect on the number of detected leaks (P < .0001).
Conclusion: This phantom study suggests that reduced radiation exposure is achievable with CT angiography at 80 kVp in small and intermediate-sized patients and at 100 kVp in large patients, with no increased risk of missing significant endoleaks measuring 6 mm or larger.