Optimal post‐contrast timing of breast MR image acquisition for architectural feature analysis

Abstract

PURPOSE: To define a post-contrast imaging time span during which diagnostic accuracy of breast magnetic resonance (MR) architectural feature analysis is maintained. MATERIALS AND METHODS: Seventy-five patients with mammographically-visible or palpable findings underwent MR examination. Three sequential post-contrast, fat-saturated, three-dimensional gradient-echo imaging runs were acquired spanning 0-90, 90-180, and 180-270 seconds after contrast injection. Five readers independently predicted the malignant potential of the MR abnormalities. RESULTS: Receiver-operator characteristics (ROC) curves were our primary measure of diagnostic accuracy. The accuracy of four readers was unchanged over the three post-contrast runs. One reader was slightly more accurate using the second and third runs than using the first. CONCLUSION: For most readers, a single post-contrast run performed at any point during the first four minutes and 30 seconds following injection should yield an equivalent diagnostic accuracy. If any time period is less optimal, it is that of our first run, performed between 0-90 seconds after contrast injection.