Does higher field strength translate into better diagnostic accuracy? A prospective comparison of breast MRI at 3 and 1.5 Tesla

Objective We performed a prospective monocentric study to inter-individually compare the diagnostic accuracy of breast MRI at 1.5 T and 3 T. Methods During a consecutive period of 12 months all 982 patients receiving breast MRI according to standard indications (problem solving, preoperative staging) were randomized to one 3 T or 1.5 T scanner both equipped with dedicated 16-channel breast coils. Protocols at 1.5 T and 3 T were optimized and in line with international recommendations. Due to restricted time slot availabilities, the randomization-key was defined as 1/10 (3 T/1.5 T). All examinations were read by two experts in breast MRI (>25 and 8 years of experience) who assigned a BI-RADS category per breast. Histopathological verification or long-term MRI follow-up (>24 months) served as standard of reference. Results were analyzed using cross tabulations, standard estimates of diagnostic accuracy, Chi-square test and Mann-Whitney U test (alpha = 5%). Results 1961 breasts were included. 1746 (89%) were imaged at 1.5 T and 215 at 3 T (11%). The 1.5 T and 3 T study cohort did not show significant differences regarding patient age (P = 0.71), results of the reference standard (P = 0.09) and indication for MRI (P = 0.53). Overall Sensitivity (94.7%), Specificity (91.4%), Accuracy (91.9%) and Negative Predictive Value (99.0%) were within the range of the literature. Pairwise comparison of Sensitivity (1,5 T/3 T: 94.1/97.9%), Specificity (91.6/%89.3%), Accuracy (92.0%/91.2%) and Negative Predictive Value (98.9%/99.3%) were without significant differences (P = 0.29-0.74). Conclusion In this prospective monocentric study, we identified comparably high diagnostic accuracy for both 1.5 T and 3 T breast MRI. Both 1.5 T and 3 T are equally suited for breast imaging.