COMPARISON OF TRANSVERSE COMPUTED TOMOGRAPHIC EXCRETORY UROGRAPHY IMAGES AND MAXIMUM INTENSITY PROJECTION IMAGES FOR DIAGNOSING ECTOPIC URETERS IN DOGS.

Abstract

Computed tomographic maximum intensity projection (MIP) images have been shown to improve reader confidence in their diagnoses and to improve detection of vascular structures and pulmonary nodules. The objectives of this method comparison study were to compare transverse source computed tomographic excretory urography (CTEU) images to two, five, and 10 slab thick MIP images for diagnosing canine ectopic ureters, compare reader confidence, and evaluate interobserver agreement. Two board-certified veterinary radiologists and a board-certified small animal internist blindly reviewed transverse source CTEU and two, five, and 10 slab thick MIP images of 24 dogs enrolled in the study. Inclusion criteria included a diagnostic CTEU and either cystoscopic or surgical confirmation of the presence or absence of ureteral ectopia. Eleven dogs were confirmed to have 17 ectopic ureters at surgery and/or cystoscopically. There was no significant difference in reader diagnoses between viewing methods or between viewing methods and the surgical/cystoscopic findings (P < 0.001). Reader confidence was significantly greater on two (P = 0.0080) and five (P = 0.0009) slab thick MIP images with significant interobserver agreement between readers for all viewing methods (P values ranging between 0.0363 and <0.001). In addition, the diagnostic accuracy of MIP images was similar to or better than transverse source CTEU images, when assessed by a radiologist. The study results suggest that CTEU is a reliable imaging technique for diagnosing canine ectopic ureters among specialists of varied experience. In addition, thin slice reconstructed MIP images improve reader confidence and potentially diagnostic accuracy, and thus their use should be considered, especially in more challenging cases.