Detection accuracy of various display modalities in CT angiography of renal artery stenosis.

Purpose: Purpose of the study was to compare the sensitivity and specificit y of various display modalities in the detection of renal artery stenosis. In particular, the difference between hard-copy reading and interactive ana lysis at the workstation was assessed. Materials and Methods: Selected pati ents (n=31) with expected suboptimal conditions for CT angiography due to l ong-standing hypertension and compromised renal function were included. Six radiologists evaluated independently a total of 77 renal arteries with 49 renal artery stenoses proven by angiography. Image analysis included: mode A: interactive display, analysis of multiplanar reformats and axial section s at the workstation, mode B: visualization of MIPs in the coronal and axia l planes as hard copies only, mode C: visualization of MIPs and axial secti ons as hard copies only. Results: The following sensitivities and specifici ties were found for internal readers and (readers from outside institutions ): mode A: 94.8%, 87.9%, mode B: 97.7% (95.1%), 80.3% (75.4%): mode C: 97.0 % (95.3%), 78.8% (76.6%). The differences were statistically not significan t (p > 0.05). Conclusions: Standardized hard copies of MIPs plus axial CT p rovide sufficient accuracy to detect renal artery stenosis compared to inte ractive imaging even in this highly selected group of patients.