pressure indicated veno-occlusive dysfunction. RESULTS• Forty-five of the 55 patients were diagnosed with corporal veno-occlusive dysfunction. 3D-CT cavernosography revealed drainage veins in all 45 cases, including cavernous veins, dorsal veins, crural veins and other emissary veins.• Compared with 3D-CT cavernosography, observing cavernous veins and the proximal part of the deep dorsal veins using MIP imaging was especially difficult because the origins of the penile veins are often behind the pelvic bone or cavernous body.• Of the patients who seemingly had leakage via the deep dorsal vein, 80.6% did not in fact have leakage via this vein, but had other leakages. The image resolution of 3D-CT cavernosography was significantly higher than that of MIP. CONCLUSION• 3D-CT cavernosography can provide highresolution images of venous drainage from any angle. We conclude that the images obtained by 3D-CT cavernosography are very helpful for both the diagnosis of corporal veno-occlusive dysfunction and the anatomical study of the human penile venous system. What's known on the subject? and What does the study add? Venous ligation surgery has been conducted as a symptomatic treatment, but the effective rate of this surgery was insufficient. We thought that one of the reasons for the low effective rate of this surgery was insufficient for the diagnostic modality. We confirmed that 3D cavernosography was, in comparison with conventional cavernosography, higher in diagnosis precision. OBJECTIVE• To examine the feasibility of threedimensional (3D) CT cavernosography in the diagnosis of corporal veno-occlusive dysfunction. PATIENTS AND METHODS• The subjects were 55 patients who had failed to respond to phosphodiesterase type 5 inhibitors. We performed pharmacodynamic infusion cavernosometry and cavernosography, using 60 mg papaverine hydrochloride.• Cavernosography was performed at 90 mmHg intracavernous pressure, using a multi-slice CT scan system. The 3D images were reconstructed using AQUARIUS NET STATION , ver.2 computer software.• For comparison with conventional cavernosography, maximum intensity projection (MIP) images were used. A flow of 20 mL/min or being more capable of maintaining 90 mmHg of intracavernous

basic type), in which the internal pudendal artery (IPA) originates from the anterior trunk at the level between the linea terminalis and the major ischial notch; Type 2, the IPA originates from the anterior trunk of the IIA at the level of the major ischial notch or more distally; Type 3, the IPA originates directly from the IIA at a level proximal to the linea terminalis; Type 4, the IPA originates together with the superior and inferior gluteal artery within 1 cm of each other; and Type 5, the penile blood supply is dependent on arteries other than the IPA, such as the obturator artery. RESULTSAmong the 290 units, eight could not be classified due to poor image quality. There were no statistically significant differences in blood flow parameters among the types of IIAs, but there was a statistically significant difference in the IPA type at the age of onset of ED. Type 1 (153 units or 53%) anatomy, was more common in patients who developed ED at an advanced age. Types 2, 3 and 4 were more common in patients with onset of ED at an early age (log-rank test P < 0.001, Study Type -Prognosis (outcomes research) Level of Evidence 2c OBJECTIVETo investigate the relationship between variations of the pelvic artery arrangement and the age at erectile dysfunction (ED) onset, as some men develop ED while relatively young, while others maintain erectile function into old age despite having cardiovascular diseases, thus congenital factors might be involved. PATIENTS AND METHODSWe examined 290 units of internal iliac arteries (IIA) in 145 patients showing repeated incomplete erectile response to intracavernosal injections with prostaglandin E 1 . Patients with cardiovascular risk factors, neurological disease or pelvic injury were excluded. The pelvic artery arrangement, evaluated by three-dimensional computed tomographic angiography, was classified anatomically into five types: Type 1 (normal or