DCE MRI is useful in evaluating breast cancer. Precise assessment of size and disease extent is required to optimize treatment plan. This can be achieved by multi-planar reconstructions or 3-D image display that requires high resolution (HR)- DCE MRI. Guideline of European Society of Breast Imaging recommended spatial resolution of at least 1.0 x 1.0 x 2.5 mm in breast DCE-MRI¹. Breast cancers can be visualized without using contrast agent as high intensity area on diffusion weighted MRI (DWI). Use of DWI in a clinical setting is limited because of image distortion due to single-shot echo planar imaging and limited spatial resolution /Signal to noise ratio (SNR). Readout segmented EPI (rs-EPI) is a recently-developed technique to obtain high resolution DW-MRI with less distortion and its value in diagnosing breast lesions has been reported ^{2,3, 4}. Another emerging technique is computed DWI (cDWI), that is a technique of synthesizing arbitrary b value diffusion-weighted images (DWI) from a set of measured b value images and applied to prostate cancer with better contrast to noise ratio CNR) ⁵ . By using these technique,HR-DWI can be an alternative of HR-DCE MRI in breast cancer without contrast agent. This study aimed to evaluate the visibility of HR- DWI using rs-EPI and cDWI calculated from the same image by comparing qualitative lesion conspicuity and maximum lesion diameter, using HR-DCEMRI as a reference standard.

Study population: 18 breast lesions with high suspicion of malignancy who underwent breast MRI. Image acquisition: HR-DWI using (TR / TE =10500 /52 sec) with a resolution of 1.1 x 1.1 x 1.4 mm and b value of 0 (NEX=1) and 850 (NEX=2) sec/mm2 were obtained using readout-segmented echo-planar imaging (rs-EPI). Based on the HR-DWI data, cDWI with hypothetical b value of 1200 sec/mm2 were calculated per pixel-based. HR-DCE with 3D-VIBE (TR /TE =4.01/ 1.63 sec, Flip Angle 15 degree, NEX=1) with a resolution of 0.7 x 0.6 x 0.8 mm are obtained during 2-5 minutes after the administration of contrast agent.

Image analysis: Lesion conspicuity was scored in 5 categories, ranging from score 5 as excellent depiction of lesion to score 1 as lesion not visible/ identified at all. Maximum diameter of the lesion was measured on the viewer with the three image series side by side. Diameter was not measured if the lesion was not visible at all (i.e. lesion conspicuity score 1). Maximum diameter of HR-DWI and cDWI was compared to HR-DCE MRI using Wilcoxon signed rank test.

Lesion conspicuity: For masses (n=16) mean scores of lesion conspicuity were 4.9, 4.1, and 3.7 for HR-DCE, HR-DWI with rs-EPI, cDWI respectively. Detailed morphology with almost no distortion can be achieved using HR-DWI with rs-EPI. The scores for NME (n=9) were 4.2, 2.3, and 2.0 for HR-DCE, HR-DWI, cDWI respectively. NME was often poorly visualized on HR-DWI and cDWI. There are four NME cases with lesion conspicuity score of 1 - i.e. not visible at all, while visible on HR-DCE MRI. Size: Maximum diameter of the lesions measured on the three different images are shown in figure 1 (Mass) and figure 3 (NME). For mass lesions, mean value of maximum diameter measured on HR- DCE, HR-DWI, and cDWI were 20.6, 19,4 and 18.7 mm respectively. Maximum diameter measured on HR-DWI was similar but slightly shorter than that on HR-DCE. The diameter on cDWI is significantly shorter compared to HR-DCE (p=0.004, Wilcoxon signed-rank test). A case of invasive cancer is shown in figure 2. For NME, mean value of maximum diameter measured on HR- DCE, HR-DWI, and cDWI were 34.9, 21.8 and 23.0 mm respectively. Typical appearance of NME associated with a mass is shown in figure 4. HR-DWI with rs-EPI can visualize mass lesion with similar or slightly shorter lesion diameter. Maximum diameter of the lesion is a key parameter used for RECIST criteria in evaluating treatment response of chemotherapy or endocrine therapy. The capacity of HR-DWI in obtaining reasonably similar size with HR-DCE MRI supports the potential use of HR-DWI in evaluating disease extent preoperatively or treatment response in breast cancer patient, even if gadolinium agent is contraindicated. In contrast, visualization of NME using HR-DWI and cDWI is not satisfactory.Masses showed relatively good lesion conspicuity on DWI with rs-EPI with similar lesion size compared to HR-DCE. The current results suggest promising results of rs-EPI as an alternative to standard DCE in evaluation breast mass lesions when contrast agent is contraindicated. Evaluation of NME using DWI with rs-EPI DWI is still challenging.