Synopsis

The low resolution of diffusion-weighted imaging (DWI) is the main limiting factor in screening breast MRI. Recently a method was proposed for one millimeter isotropic DWI covering the entire breasts. Based on a readout-segmented DW-EPI sequence, 3 stacks of thick slices (3 mm) are acquired with 1mm-shifts in the slice direction; then a 1x1x1 mm³ dataset is obtained using a super-resolution reconstruction (SRR). In this study we further validate the method using a commercial diffusion breast phantom, in terms of SNR, resolution and ADC values. Additionally, SMS acceleration is investigated to reduce the total scan time below 10 min.

Introduction

Methods

Data acquisition

Breast MRI examinations were performed using a 3T clinical MR scanner (MAGNETOM Prisma; Siemens Healthcare, Erlangen, Germany), with dedicated 18-channel breast array coil. The unaccelerated SRR protocol (without SMS) was comprised of three sets of rs-EPI low resolution (LR) axial images aquired with 1 mm shifts at each repetition (whole breasts coverage, 1x1x3 mm³, b = 0, 200, and 800 s/mm²; 50 slices, TR/TE=10410/56ms, FOV 320x160mm²). It was applied to eight healthy subjects and a breast diffusion phantom (High Precision Devices Inc, Boulder, USA). The SMS accelerated protocol was set with a prototype SMS-rs-EPI. Parameters were the same, except slice acceleration=2, TR/TE=5440/60ms (TR had to be shortened since the number of excited slices is halved with SMS), and FOV 320x162mm². It was applied to one healthy subject and the breast diffusion phantom.

Reconstruction and data analysis

Reconstruction of isotropic volumes (1x1x1 mm³) was performed for both rs-EPI and SMS-rs-EPI using SRR with Beltrami regularization (3). ADC maps were calculated using the 3 b values, by linear fitting the logarithmic relative intensity against b₀. In the phantom, six regions of interest (ROI) were drawn to extract the different compartments with known ADC. Furthermore, spatial resolution was assessed visually using the cylinder compartments with varying radii. Finally, SNR was calculated using the double acquisition method (using the sum and subtraction of two SRR images), with an ROI drawn in a homogeneous area (i.e. with small variation in noise level) (5):

$$SNR = \frac{Signal_{ mean ROI}}{\sigma_{ROI}}$$

$$Signal_{ mean} = \frac{mean (S_1+S_2)_{ROI}}{2}$$

$$\sigma_{ mean} = \frac{\sigma(S_1-S_2)_{ROI}}{\sqrt{2}}$$

In the subjects, ADC values were obtained from carefully drawn ROIs in left and right breast, selecting a region with no fatty tissue and showing a texture characteristic of normal fibroglandular tissues.

Results

The total acquisition time of three acquisitions with/without SMS acceleration was 9:36 min/15:42 min.

Figure 1 shows the reconstructed phantom data, including the spatial resolution patterns and the diffusion compartments with varying ADC values. Table 1 shows the theoretical and measured ADC values in the phantom with/without SMS. The difference in ADC values was 1.9% between the unaccelerated SRR protocol and its SMS-accelerated version. Table 2 shows signal, noise and SNR with the two protocols. A minor reduction of SNR is obtained with SMS.

Figure 2 shows an example isotropic 1x1x1 mm³ ADC map in a subject, with/without SMS. Table 3 shows the mean ADC values (averaged over the eight subjects and the two breasts) which were 1.91 x10^-3 mm²/s ± 0.29 with unaccelerated SRR. In the subject with/without SMS the ADC values were 2.05 and 2.1 *10^-3 mm²/s.

Discussion and Conclusion

Acknowledgements

Grant funding: ANR-17-CE19-0022 (Bracoil project) and INCA, Plan Cancer 2014-19 (Screenbreast project). The authors also thank Khalid Ambarki (Siemens Healthineers, Saint-Denis, France) and Wei Liu (Siemens Healthineers, Shenzen, China) for providing the SMS-rs-EPI sequence prototype, and Institut de Cancérologie de Lorraine, INSERM, FEDER and Région Lorraine.

References

1. Kul S, Metin Y, Kul M, Metin N, Eyuboglu I, Ozdemir O. Assessment of breast mass morphology with diffusion-weighted MRI: Beyond apparent diffusion coefficient: Breast DWI for the Evaluation of Mass Morphology. J Magn Reson Imaging [Internet]. 2018 May 7 [cited 2018 May 15]; Available from: http://doi.wiley.com/10.1002/jmri.26175

2. Filli L, Ghafoor S, Kenkel D, Liu W, Weiland E, Andreisek G, et al. Simultaneous multi-slice readout-segmented echo planar imaging for accelerated diffusion-weighted imaging of the breast. Eur J Radiol. 2016 Jan;85(1):274–8.

3. Maya Delbany, Bustin A, POUJOL J, Thomassin-Naggara I, Felblinger J, Vuissoz P-A, et al. One-millimeter isotropic breast DWI: evaluation of a super-resolution strategy in terms of signal-to-noise ratio, sharpness and apparent diffusion coefficient. Magn Reson Med. In press;

4. Setsompop K, Gagoski BA, Polimeni JR, Witzel T, Wedeen VJ, Wald LL. Blipped-controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced g-factor penalty. Magn Reson Med. 2012 May 1;67(5):1210–24.

5. Reeder SB, Wintersperger BJ, Dietrich O, Lanz T, Greiser A, Reiser MF, et al. Practical approaches to the evaluation of signal-to-noise ratio performance with parallel imaging: Application with cardiac imaging and a 32-channel cardiac coil. Magn Reson Med. 2005 Sep;54(3):748–54.