Comparison of the diagnostic value of simultaneous multislice (SMS) accelerated diffusion-weighted echo planar imaging (EPI) of malignant and benign lesions of the breast compared with a reference EPI sequence.
The study was approved by the institutional ethics committee. Sixty-eight patients were examined with a diffusion-weighted EPI (reference EPI; TE = 54 milliseconds; TR = 9000 milliseconds; TA, 3:27 minutes) and a diffusion-weighted SMS accelerated EPI (SMS EPI; acceleration factor 2; TE = 58 milliseconds; TR = 4300 milliseconds; TA, 1:53 minutes) in addition to the standard magnetic resonance imaging (MRI) protocol. Further acquisition parameters were as follows: 3 T MAGNETOM Skyra (Siemens Healthcare, Erlangen, Germany), 2.5-mm isotropic resolution, field of view = 185 to 190 × 350 mm2, 62 slices, b = 50 and 800 s/mm2 with 1 and 4 averages, respectively. A dedicated 16-channel bilateral breast coil was used for imaging. Image quality was evaluated with respect to the presence of artifacts, signal voids, and quality of fat suppression. These parameters were rated using a 5-point Likert scale (1 = very strong to 5 = negligible). The apparent diffusion coefficient (ADC) was measured in 72 focal lesions (46 breast carcinomas and 26 benign lesions), and the diagnostic value of the 2 datasets was statistically evaluated and compared. The evaluation was performed a second time excluding cysts.
Artifacts and signal voids were negligible in both sequences (mean on Likert scale for reference EPI 4.68 vs SMS EPI 4.65, P = 0.52, and mean on Likert scale for reference EPI 4.85 vs SMS EPI 4.77, P = 0.14). Fat suppression was significantly better in SMS EPI (mean on Likert scale 3.28 vs 2.97, P < 0.001, Pearson r = 0.49). For benign lesions, the mean ADC in both EPI sequences was 1.86 · 10−3 mm2/s. For malignant lesions, a mean ADC of 0.90 · 10−3 mm2/s for the reference EPI and 0.89 · 10−3 mm2/s for the SMS EPI was found. No significant difference between the EPI sequences was observed for ADC values (P = 0.75) and for the area under the curve (SMS, 0.985; no SMS, 0.975). The cutoff for differentiation of benign and malignant lesions was at ADC = 1.42 · 10−3 mm2/s for SMS EPI (sensitivity, 1; specificity, 0.88) and at 1.23 · 10−3 mm2/s for the reference EPI (sensitivity, 1; specificity, 0.92). Excluding the cysts, the cutoff for differentiation of benign and malignant lesions was at ADC = 1.11 · 10−3 mm2/s for SMS EPI (sensitivity, 0.89; specificity, 0.93) and at 1.23 · 10−3 mm2/s for the reference EPI (sensitivity, 1; specificity, 0.87).
Our data indicate that SMS acceleration can be used for diffusion imaging in breast MRI in clinical practice. Simultaneous multislice EPI achieved the same diagnostic accuracy in breast MRI, but in a substantially reduced scan time.
From the *Institute of Radiology, University Hospital Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg; and
†Siemens Healthcare, Erlangen, Germany.
Received for publication December 21, 2018; and accepted for publication, after revision, February 2, 2019.
Conflicts of interest and sources of funding: This study was supported by the Research Foundation Medicine at the University Clinic Erlangen, Germany. Funding for the positions of F.B.L. and T.P. by the Deutsche Forschungsgemeinschaft (LA 2804/6-1) is gratefully acknowledged.
Correspondence to: Sabine Ohlmeyer, MD, Institute of Radiology, University Hospital Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Maximiliansplatz 3, 91054 Erlangen, Germany. E-mail: firstname.lastname@example.org.
Online date: April 3, 2019