Purpose: To evaluate the diagnostic accuracy of virtual vascular endoscopy
(VVE) in the detection of accessory renal arteries.
Methods: We retrospectively reviewed the CT angiography data sets of 67 pat
ients (29 male and 38 female; age range 17-72 years, mean age 53 years) ima
ged for the study of the renal arteries, and affected by renovascular hyper
tension. All patients also had intraarterial digital subtraction angiograph
y (DSA). CT angiography data sets were processed to obtain maximum intensit
y projection (MIP) and surface-rendered VVE of the aorta. Axial images, MIP
, and VVE were evaluated separately and in combination in the detection of
accessory renal arteries. Their results in terms of sensitivity and specifi
city were then compared with DSA.
Results: Axial images had a sensitivity of 88% and specificity of 94% for a
ccessory renal artery detection, MIP had a sensitivity of 88% and specifici
ty of 98%, and VVE had a sensitivity of 63% and a specificity of 88% (p < 0
.05 vs DSA), but these increased to 88% and 98% respectively if endoscopic
views were integrated with the other display techniques.
Conclusion: VVE based on surface rendering does not add substantial benefit
s to CTY angiography; by contrast MIP is the most accurate display techniqu
e for the detection of accessory renal arteries.