The investigators will use integrated PET/MR for the goals below:

1. Use of PET-guided proton MRS (MR spectroscopy) and DCE MRI (dynamic contrast-enhanced MRI) for patients who will receive NAC (neoadjuvant chemotherapy) for breast cancer to monitor treatment response.
2. Use of dynamic and static PET to monitor treatment response for NAC, and to investigate the correlation of PET results versus MRS, DCE MRI.
3. Compare clinical staging by PET/MR and by clinical assessment.
4. On pre-chemotherapy studies, to investigate the association of molecular marker status with the dynamic and static PET, MRS and DCE MRI parameters.

Breast MRI with dynamic contrast-enhanced series (DCE MRI) is sensitive for breast cancer diagnosis (sensitivity 95-100%) with variable specificity (37-97%). Breast MRI was also used for pre-operative staging, and monitoring of therapeutic response of neoadjuvant chemotherapy (NAC). DCE MRI analysis with semi-quantitative and pharmacokinetic method can discriminate the responders versus non-responders during NAC. Analysis of choline peak on proton MR spectroscopy (MRS) can increase the MRI specificity of breast lesion diagnosis to 82-100%. Choline analysis was also used for monitoring treatment response of NAC. The change of choline integral can parallel the response status of NAC and was well correlated (r=0.91; P=0.01) with the change of lesion size, and the choline change can be found as early as 24 hours after first dose of chemotherapy.

Breast Positron Emission Tomography (PET) with 18F-FDG (2-deoxy-2-(18F)fluoro-D-glucose) was mainly used for staging and monitoring of treatment response of NAC, with the reduction of FDG uptake (P<0.001) was more evident than that of tumor size (P=0.005). However, FDG PET displayed a limited role in evaluation of primary breast cancer, brain and axillary lymph nodes metastases. There were publications regarding combined FDG PET/CT and breast MRI for breast cancer diagnosis and monitoring NAC response. The SUV on static PET and MRI findings were correlated well with molecular marker status of breast cancer (ER, PR, HER2), and were associated with clinical outcome. The change of choline integral on MRS was well correlated to the peak of SUV during NAC (r=0.84,P=0.02).The changes of dynamic PET parameters including rate constants for uptake, washout and FDG influx were moderately correlated with the DCE MRI parameters and can reflect the response status of NAC. However, PET/CT and MRI were performed by two machines at different time, so breast positioning is different, causing the lesion targeting sometimes difficult. Moreover, the selection of ROI/VOI for PET and MRS, DCE MRI is subjective with inter-observer bias. A new technology- PET/MR- can solve these problems. PET/MR has less radiation and offers more soft tissue details than PET/CT. A most recent PET/MR design- integrated PET/MR- is commercially available. Using the integrated PET/MR, patients can undergo whole body PET and MRI at the same time, followed by dedicated protocol for specific organ of primary tumor origin. The ROI/VOI of breast DCE MRI, MRS can be selected according to SUVmax site from PET, which is more objective and can ensure that the VOI/ROI is at same location across all techniques. PET/MR showed comparable reliability to PET/CT for detection of oncologic diseases, and contributed even more changes of clinical management than PET/CT. However, use of PET/MR for breast cancer was seldom reported.

The investigators will use integrated PET/MR for the studies below:

1. Use of PET-guided proton MRS and DCE MRI for patients who will receive NAC for breast cancer to monitor treatment response.
2. Use of dynamic and static PET to monitor treatment response for NAC, and to investigate the correlation of PET results versus MRS, DCE MRI.
3. Compare clinical staging by of PET/MR and by clinical assessment.
4. On pre-chemotherapy studies, to investigate the association of molecular marker status with the dynamic and static PET, MRS and DCE MRI parameters.

Inclusion Criteria:

1. Women aged 25-75 years old.
2. Women with recently diagnosed breast cancer and who will receive NAC to reduce tumor burden before surgery. (including locally advanced breast cancer (LABC) according to clinical assessment; or tumor size > 2cm, that is, at least T2 in TNM staging).

Exclusion Criteria:

1. Estimated GFR (eGFR) < 60 mL/min/1.73 m2 and blood glucose > 135 mg/dl; Past or present history of acute renal failure, renal dialysis, diabetes mellitus.
2. Women who received metallic fixation, coronary artery stent in recent 3 months; or women who received mechanical valve replacement that is not compatible with MR magnet; or women with aneurysmal clips, pacemakers.
3. Past history of claustrophobia.
4. Women who are pregnant or who are planning to be pregnant, or who are lactating (though the possibility in our target population should be very low)
5. Past history of breast cancer within recent 5 years before the currently diagnosed breast cancer.
6. Women who received chemotherapy for other disease entity in recent 1 year.
7. Women who cannot cooperate with the examinations.

undecided yet