Breast MRI: intraindividual comparative study at 1.5 and 3.0T; initial experience.

PURPOSE: To prospectively and intraindividually compare breast magnetic resonance imaging (MRI) at 1.5 Tesla (T) and 3.0T. METHODS: A prospective intraindividual Ethics Committee- approved study was performed in 31 women (average age 58.6±12.3 years), with 114 lesions (9 benign, 105 breast cancers; 24 patients with unilateral and 7 with bilateral cancers). Axial bilateral breast high-spatial resolution contrast-enhanced dynamic MRI was performed at 1.5T using 3 dimensional (3D) dynamic gradient-echo sequences in all patients (spatial resolution 1.1×0.7×2 mm; temporal resolution 41 sec per dynamic acquisition), and after 24-48 h at 3.0T (0.6×0.6×1.7 mm; temporal resolution 65 and 72 sec per dynamic acquisition). Contrast enhancement ratio, number and features of enhancing lesions, image quality and reliability were compared by two radiologists independently. RESULTS: 102 cancer lesions were detected at 1.5T and 105 cancer lesions were detected in 31 patients at 3.0T. One cancer lesion was observed at 1.5T which was missed at 3.0T, and 3 cancer lesions and one high-risk lesion (LCIS) were detected at 3.0T while missed at 1.5T. Enhancement rates were significantly higher at 1.5T (224.5±100.2) compared to 3.0T (133.7±38.3). Better image quality was observed at 3.0T. Interobserver reliability was higher at 3.0T (p= 0.684) compared to 1.5T (p= 0.351). CONCLUSION: Detection of breast cancer shows a trend of better performance at 3.0T than at 1.5T.