1) To describe the technical elements needed to perform high-quality breast MRI. 2) To describe and illustrate the pulse sequences needed for high-quality breast MRI. 3) To describe and illustrate the importance of simultaneously achieving high in-plane spatial resolution, thin slices, adequate temporal resolution, adequate signal-to-noise ratios, and full coverage of both breasts in breast MRI. 4) To show examples of high-quality and sub-standard breast MRI exams.

To determine the diagnostic performance of diffusion kurtosis imaging (DKI) of the breast lesions for the detection of malignant breast tumors.

IRB approval and written informed consent was obtained. During a 10-months period, we prospectively evaluated 51 patients (mean age, 52.0 years; range, 13-86 years) with 58 breast lesions including 21 benign lesions and 37 malignant tumors. DKI was performed with a single-shot echo-plannar sequence with multiple b values (0, 100, 500, 1000, 1500, and 2000 sec/mm2). We computed the mean kurtosis (MK) and apparent diffusion coefficient (ADC) (10-3 mm2/s) over regions of interest encompassing the entire tumor using diffusion kurtosis model programed by MATLAB software (Mathworks, Natick, Mass). The diagnostic performance of MK and ADC value for the detection of malignant breast tumors were compared.

MR image acquisition and analysis were successful in all our study patients. MK was significantly higher in malignant tumors (1.13 ± 0.28) than in benign lesions (0.74 ± 0.22) (P < 0.001). Mean ADC value was significantly lower in malignant tumors (1.17 ± 0.35) than in benign lesions (1.72 ± 0.41) (P < 0.001). For the detection of malignant tumors, there was no significant difference in AUC between MK and ADC value, whereas ADC value (71.4%) had a greater sensitivity than MK (62.2%) (P = 0.007) and MK (100%) had a greater specificity than ADC value (86.5%) (P = 0.04).

DKI was feasible in breast MRI. Our study findings suggest that a combination of MK and ADC may provide the additional value for the detection of malignant breast tumors.

When MRI is performed to evaluate the breast lesions, diffusion kurtosis imaging may improve the diagnostic confidence of lesion characterization in addition to conventional diffusion imaging analysis in breast MR imaging.

To compare mono-exponential model DWI and intravoxel incoherent motion(IVIM) DWI in characterizing different subtype and different grade of breast lesions.

51 malignant and 47 benign breast lesions in 93 patients were performed with mono-exponential DWI(b =0,600s/mm2) and bi-exponential analysisDWI (b =0,50,100,150,200,500,800 s/mm2) at 3.0T MRI. Apparent diffusion coefficient (ADC), as well as IVIM-based parameters:true diffusion coefficient (D), perfusion fraction (f), and pseudo-diffusion coefficient (D*) were compared among different subtype and different grade lesions . Receiver operating characteristic (ROC) was used for lesion discrimination.

All the data were fitted well (R2>0.90). In the lesions except cyst, D value was significant lower than ADC (p<0.05).ADC, D and D* in malignant tumors were significantly smaller than those of benign ones (P = 0.000 for ADC and D ,0.002 for D*, respectively), f value was higher than that of benign lesions (P=0.000). D value showed increasing order as the following: Invasive ductal carcinoma (IDC)< ductal carcinoma in situ (DCIS) ˜Intraductal papilloma (IDP) <Fibroadenoma (FA)< breast cyst. ROC demonstrated D had the best performance in identification of malignancy from benign lesion (AUC= 0.945) and discriminating DCIS from IDC (AUC= 0.791) than that of ADC,f and D*. Furthermore,a direct inverse correlation had been observed between D value and histological IDC grade,while no other parematrers showed difference among varied grade of IDC (P>0.05).

IVIM-DWI provides quantitative measurement of cellularity and vascularity properties within breast lesions and D showed better diagnostic ability in discrimination malignancy and tumor grading than ADC. therefore IVIM are expected to enhance the role of MRI in diagnosis, monitoring, and treatment screening of breast cancer.

IVIM provides noninvasive sensitivity to lesions microenvironment properties and has the potential to improve the specificity of breast MRI without contrast agent ,thus it is recommended as part of the MR exam for screening high-risk women.

To correlate the enhancement parameters of dynamic contrast-enhanced magnetic resonance imaging (MRI) with prognostic factors and immunohistochemical subtypes of breast cancer.

A total of 81 breast carcinomas were included in our study. We obtained the following enhancement parameters: 1) background parenchymal enhancement (BPE) and BPE coefficients (BEC) from bilateral breasts, 2) the number of vessels per breast as a representation of whole-breast vascularity. In 50 patients, 3) semiquantitative parameters of tumors (the initial enhancement percentage, the peak enhancement percentage, the time to peak enhancement, the signal enhancement ratio) and 4) perfusion parameters (Ktrans, kep, ve and iAUC) from tumors and ipsilateral breasts. Correlations among parameters and prognostic factors, including tumor size, axillary node status, nuclear grade, histologic grade, estrogen receptor (ER) expression, progesterone receptor (PR) expression, Ki-67, human epidermal growth factor receptor 2 (HER-2) expression, epidermal growth factor receptor (EGFR) expression, bcl-2, CK5/6 and subtypes categorized as luminal, triple negative and HER-2 were analyzed.

BPE was significantly correlated with EGFR expression (p=0.040). BEC was significantly higher in tumors larger than 2 cm than in tumors smaller than 2 cm (p=0.001). The vessel numbers in ipsilateral breasts were higher in tumors larger than 2 cm than in tumors smaller than 2 cm (p=0.034), with higher nuclear grades (grade 3) than with lower nuclear grades (grade 1,2) (p=0.001) and with PR-negative rather than with PR-positive (p=0.010) results. The mean Ktrans was higher in Ki-67-positive tumors than Ki-67 negative tumors (p=0.002). The mean kep was higher in Ki-67-positive tumors than in Ki-67-negative tumors (p=0.005) and in CK5/6-positive tumors than in CK5/6-negative tumors (p=0.015). The mean Ktrans was lower in the ipsilateral breast parenchyma with HER-2-positive tumors compared to HER-2-negative tumors (p=0.012).

The BPE, BEC and ipsilateral whole-breast vascularity, higher Ktrans and kep of breast cancer and lower Ktrans and iAUC of ipsilateral breast parenchyma may serve as additional predictors of poor prognosis of breast cancer.

Enhancement parameters on breast MRI can predict the prognosis and subtypes of breast cancer and is recommended for the preoperative evaluation of breast cancer patients.

Detection of breast cancer at earlier stages has raised concern of overtreatment in subgroups of patients, while treatment failure still occurs in other. Continuing need exists for prognostic models tailored to individual patients at time of diagnosis. Preoperative core biopsy results in discordant assessment of tumor grade in up to 40% compared to postoperative assessment. Imaging features may potentially close this gap, as they provide full overview of the tumor. Aim of this study is to assess the potential of biomarkers at 7T functional Breast MRI to characterize the proliferative nature of breast tumors in-vivo.

A one-stop-shop protocol was developed at 7T MRI (Philips) including high temporal (HT; 5 s; 2,8mm³ isotropic) and high spatial (HS; 90 s; 0,7mm³-1,0mm³ isotropic) dynamic contrast-enhanced (DCE) magnetic resonance (MR) imaging, diffusion-weighted imaging (DWI), and phosphorus spectroscopy (31P-MRS) to analyze tumor metabolism. Sixteen women (age 53-70 years) with histologically proven invasive breast carcinoma on biopsy were scanned prior to surgery. DCE characteristics were assessed according to BI-RADS. ADC-values were calculated and hypointense tumor areas scored. Localized 31P-MR spectra were assessed and scored (1-5) based on degree of tumor proliferation using a newly developed lexicon. Tumor characteristics on pathology were assessed from the resection specimen and correlated to the MRI features. Explorative analyses were performed using box plots, Pearson Chi-Square and Krusal Wallis tests.

In 16 patients, 18 malignant lesions were detected on HS DCE-MRI. The mean largest tumor diameter was 22mm (range 8-58). Time to enhancement on HT DCE-MRI ranged from 12s to 29s. Shortest interval was observed in a rim-enhancing triple-negative tumor. First observations showed correlations between the 31P-MRS score and mitotic cell index (N=11; p=0,02) as well as a trend between ADC and modified Bloom-Richardson tumor grade (N=11; p=0,097).

A one-stop-shop imaging protocol for breast MRI at 7T was developed to explore prognostic and predictive tumor biomarkers in-vivo. First explorations indicate feasibility to visualize tumor grade in-vivo.

Imaging of breast cancer biomarkers in-vivo using a one-stop-shop 7T Breast MR imaging protocol allows stratification of tumor proliferation, an important predictive marker used in therapy selection.

1) Understand the physical basis of diffusion imaging and methods used to acquire diffusion-weighted data. 2) Understand the clinical applications of diffusion-weighted imaging for cancer diagnosis and assessment of response to therapy. 3) Be familiar with the challenges of breast diffusion imaging and technical considerations for protocol optimization. 4) Future directions.

Percutaneous biopsy is mandatory for all suspicious BMRI-detected lesions (BI-RADS 4 of the Breast Imaging Reporting and Data System), but the malignancy rate is variable (from 2 to 95%) and BMRI-guided biopsy is an expensive procedure, frequently resulting in benign histopathology. Our purpose was to investigate whether DWI and tUS could help in this setting by reducing the number of cases assigned as BI-RADS 4.

From January 2008 to December 2012, 1757 patients underwent BMRI (1.5T) including T2-weighted sequences, DWI (b-values: 0-900s/mm2) and dynamic study. A BI-RADS score was assigned according to conventional morpho-dynamic criteria. For each suspicious enhancement (BI-RADS 4), the Apparent Diffusion Coefficient (ADC) value was quantified and an ultrasonographic correlate was searched. When both DWI and tUS downgraded the suspicious enhancement to BI-RADS 3 (probably benign), a short term follow-up was recommended. If at least one of the two methods (DWI and/or tUS) confirmed the suspicion, the lesion remained BI-RADS 4 and a biopsy was done using MRI or US-guidance. Histopathological results and radiological follow up data (minimum 6 months) were recorded and considered the gold standard. Diagnostic performance indicators of MRI+DWI and tUS were compared using Chi-square test.

BMRI detected 152 BI-RADS 4 lesions in 120 patients: 52% (90/152) of them were proven to be breast cancers. Eighty percent (121/152) of the lesions were also DWI visible. The sensitivity, specificity, positive and negative predictive value of DWI in the characterization of these lesions were 98%, 87%, 93 and 96%, respectively. An ultrasonographic correlate was found in 66% of MRI lesions: the sensitivity, specificity, positive and negative predictive value of tUS were inferior to those of DWI (p<0,05) (77%, 82%, 86% and 74%, respectively). Twenty-eight percent (43/152) of the suspicious lesions were downstaged to BI-RADS 3 by means of both DWI and tUS: 7% of these revealed to be malignant during subsequent follow-up (2 DCIS and 1 tubular breast cancer).

The combination of Quantitative DWI and tUS showed a high accuracy in the characterization of BMRI-detected suspicious lesions, resulting in a reduction of false positives.

The use of DWI and tUS could theoretically reduce the need of unnecessary preoperative breast biopsies in case of BI-RADS 4 enhancing lesions.

We compared measurements of apparent diffusion coefficient (ADC) within the whole breast tumor vs. a small intratumoral region of interest (ROI) to differentiate malignant from benign tumors and assessed whether the ADC parameters represent surrogate markers for tumor prognostic characteristics.

Approval of the institutional ethics board and institutional approval were obtained. The study protocol included 3.0T structural breast MRI and diffusion weighted imaging (DWI). Patients were selected according to the recommendations of EUSOMA and the local guidelines. Forty-nine patients (mean age, 59±12 years; range, 36-82 years) and 49 lesions (17 benign, 32 malignant) were included to this prospective study. Two observers measured the ADC values (mean, standard deviation, kurtosis and skewness) from a) the whole lesion (WL-ROI), avoiding cystic, hemorrhagic and necrotic regions b) six small intratumoral ROIs (SI-ROIs) with lowest-appearing values on ADC map. WL-ROI and SI-ROI with second-lowermost mean value were used for statistical analysis. Data on estrogen and progesterone receptors, HER2-overexpression, tumor grade, Ki-67, vascular and lymph duct invasion and metastasis to axillary lymph nodes were collected.

Using SI-ROI ADC mean values, reproducibility of the measurements proved to be excellent (κ=0.75; Intra-Class Correlation Coefficient, 0.904). In receiver operating characteristic curve analysis, area under the curve was 0.891 for observer 1 and 0.881 for observer 2. Using cut-off value of 0.602×10-3 mm2/s, sensitivity of 82.4%, specificity of 87.5% and overall accuracy (OA) of 85.7% were reached for diagnosing malignant lesions. ADC measurements from SI-ROIs proved to be more accurate than WL-ROIs (OA=67.3%, P<0.05). SI-ROI ADC values inversely correlated to the presence of axillary metastases (P=0.008), and to vascular invasion (P=0.003). There was no independent correlation between ADC values and tumor grade, estrogen, progesterone, HER2 or Ki-67 expression.

Measuring ADC values from a small intratumoral ROI is clinically more accurate than using the whole tumor ROI in assessment of breast tumors in 3.0T MRI and may help in tumor characterization.

When evaluating breast tumor MRIs, ADC measurements should be targeted to most suspicious subregion instead of the whole tumor.

To measure apparent diffusion coefficient (ADC) values in malignant lesions and evaluate their relationship with classical and molecular prognostic factors and Oncotype Dx scores.

This HIPAA compliant retrospective study consisted 212 consecutive patients with known cancers who underwent 3.0T MRI with DWI (b=0 and 600 s/mm2) between Jan' 2011and Jan' 2013. Lesions < 0.8 cm, lesions undergoing neoadjuvant chemotherapy or suboptimal DW images were excluded. ADC was analyzed on 148 malignant lesions in 135 patients. A region of interest was drawn within each lesion on DW images, avoiding any cystic/necrotic portion. Patient characteristics, classical histological prognostic factors (histologic type, grade, size, and lymph node (LN) status), molecular factors (ER, PR, and HER2) and genetic factors (BRCA, Oncotype DX scores) were reviewed and recorded. The relationships between ADC values and patient characteristics and prognostic factors were analyzed. Statistical analysis was performed using Student's t-test and ANOVA (statistical significance was established at p= 0.05). ADC values are measured in units of 10-3 mm2/s.

The mean ADC value of the 148 malignant lesions was 1.00±0.170. The ADC values in lesions were not influenced by the BPE or breast density (respectively p=0.550 and p=0.967). The mean ADC values were significantly lower for the invasive ductal carcinoma (p=0.015), mass enhancement (p<0.001), BRCA positive lesions (p=0.032) compared to DCIS, invasive lobular carcinoma, non mass enhancement and BRCA negative lesions. The mean ADC values tended to be lower in premenopause women, high grade, LN positive, triple negative lesions (though not statistically significant). No statistical significant difference was observed in the ADC values among the different subgroups in size (<2cm, 2-5cm, >5cm), and molecular prognostic factors (ER positive, HER positive, TN). According to Oncotype Dx score (available for 27/41 ER positive tumors with negative LN) ADC values were higher in low risk (0.106±0.207) than in intermediate risk tumors (0.957±0.105), even if not statically significant different (p=0.100).

Our study shows that ADC may be a potential clinical adjunct in the evaluation of prognostic factors mostly in relation to the malignant lesion aggressiveness.

ADC may be a potential clinical adjunct in the evaluation of breast cancer prognostic factors.

To determine whether preoperative MRI findings differ according to breast cancer subtype and to examine the relationship between the pattern of recurrence and breast cancer subtype in women treated with breast conserving therapy (BCT).

A total of 102 primary breast cancer patients (mean age, 45 years; range, 22-78 years) treated with BCT who had preoperative breast MRI and locoregional recurrence after BCT between September 2003 and December 2012 were included in the study. Patients who underwent neoadjuvant chemotherapy or surgical excision prior to MRI were excluded. Two breast imaging radiologists blinded to the clinicopathologic data assessed fibroglandular tissue (FGT) and background parenchymal enhancement (BPE) on MRI using BI-RADS criteria. Presence of multifocal/multicentric disease and lymph node involvement were evaluated. The pattern of recurrence and detection method were examined. Breast cancer subtypes were defined as luminal (ER+ and PR+), HER2+ (ER-, PR-, and HER2+), and triple-negative (TN; ER-, PR-, and HER2-). MRI and clinical features were compared between the breast cancer subtypes.

The 102 cases were classified as 56 (55%) luminal, 17 (17%) HER2+, and 29 (28%) TN subtype. Women with dense breasts were more likely to have luminal subtype compared to HER2+ or TN subtypes (95% vs (71%, 79%), p = 0.013). Multifocal/multicentric disease was more frequently detected by MRI in HER2+ subtype, compared to luminal or TN subtypes (59% vs (20%, 21%), p = 0.002). Ipsilateral breast cancer recurrence was more frequently observed in HER2+ subtype, compared to luminal or TN subtypes (88% vs (50%, 62%), p = 0.018). Compared to luminal subtype, HER2+ and TN subtypes were more likely to be associated with clinically detected recurrence (11% vs (41%, 41%), p = 0.002). There were no significant differences in BPE and lymph node involvement between subtypes.

Preoperative breast MRI is more likely to detect multifocal/multicentric disease in HER2+ breast cancer and FGT on MRI is more likely to be associated with luminal breast cancer. Patients with HER2+ and TN breast cancers more frequently have clinically detected recurrence.

The use of preoperative breast MRI and the postoperative imaging follow-up strategy could be tailored according to breast cancer subtype in women treated with BCT.

To determine the influence of the size and position of the ROI in Diffusion Weighted Images (DWI) of breast lesions on the Apparent Diffusion Coefficient (ADC) values and on discriminating benign from malignant lesions.

Sixty-four patients with 72 breast lesions (52 malignant and 20 benign) underwent breast DWI. ADCs were calculated for b-value pairs: 0-1000, 0-800, 0-500, 0-200 and 0-50 s/mm2. In each lesion 4 oval regions of interest (ROI) were drawn, ROI1- ROI4. ROI1 encompassed as much of the lesion as possible, while avoiding surrounding tissue, ROI2 (0.5 cm2) was located in the middle of the lesion and ROI3 (0.5 cm2) and ROI4 (1.0 cm2) were selections within the lesion yielding the lowest ADC value. ROI3 and ROI4 were compared to determine the influence of the size of the ROI. ROC analysis was used to quantify the diagnostic accuracy of the ROI methods with the different b-value pairs Statistical significance was determined with an independent sample t-test for malignant lesions and Mann-Whitney U test for all and benign lesions.

Lower b-value pairs generally showed higher ADC values in the lesions. Benign and malignant lesions significantly differ for almost every b-value pair (p<0.001). There was a significant difference between ROI3 and ROI4 for malignant lesions (p=0.005) with a higher accuracy for ROI3 (0.943 versus 0.932), probably due to reduced partial volume effect. The ADC outcomes of b-values 0-1000 and 0-800 s/mm2 met a higher specificity than the lower b-value pairs, that is, up to 70-75% for ROI1 and ROI3 when choosing a sensitivity and negative predictive value of 100%. The AUC was highest for ROI3 using b values 0-1000 and 0-800s/mm2 (0.965 and 0.964, respectively).

The size and the position of the ROI influenced the ADC values of benign and malignant breast lesions in DWI. ROI3, a small volume selected for the lowest ADC within the lesion, had the highest accuracy in differentiating benign from malignant lesions, with b-value pairs 0-1000 and 0-800 s/mm2.

Different ROI methods influence the ADC in breast DWI, therefore a ROI (0.5 cm2) positioning at the lowest ADC value within the lesion with b-value 0-1000 or 0-800s/mm2 is recommended.

To assess the value of multiparametric breast MRI (including morphology, DCE MRI and DWI with Apparent Diffusion Coefficient (ADC) mapping) at 3T in distinguishing among DCIS, Luminal A and B, HER2 positive, and Triple Negative breast cancer phenotypes

Our institutional review board approved the study. We included 219 patients with 234 lesions patients who underwent bilateral breast MRI at 3T (mean age 53+11.5 year). Both high temporal (15 sec) DCE and high spatial resolution (0.5 mm2 voxel size) MRI were acquired along with DWI with ADC mapping. Regions of interest were drawn on the ADC maps of breast lesions and normal appearing glandular tissue (GT). Morphologic features, DCE-MRI results (kinetic curve type), GT and lesion absolute and normalized ADC values were included in multivariate models for prediction of breast cancer histological subtypes. Area under ROC curve analysis was performed

Of 234 breast cancer lesions, 12% of were DCIS, 47% Luminal A, 22.2% Luminal B, 4.3% HER2 positive, and 14.5% triple negative. Lesion morphology (combining type of lesion with margin/distribution), Kinetic curve type, time to peak enhancement, and both absolute and normalized ADC values were univariate predictors of breast cancer phenotypes with an AUC 0.61-0.79. Combining lesion volume, morphology, kinetic curve type, internal enhancement, and normalized ADC value showed the best accuracy in predicting estrogen receptor expression, while combining lesion diameter, morphology and ADC value showed the best diagnostic accuracy in predicting progesterone receptors expression, and combining lesion diameter, morphology, and normalized ADC value showed the best accuracy in predicting the HER2 receptor expression. For the phenotypes characterization, the multivariate diagnostic model combining lesion morphology, kinetic curve type, and normalized ADC value showed the best diagnostic accuracy (AUC 0.83)

Multiparametric MRI including morphology, DCE and DWI can characterize breast cancer phenotypes with a very good diagnostic accuracy (AUC =0.83) at 3T

Breast cancer tumors with the same histological characteristic may carry different prognosis and response to treatment due to the difference at the molecular level. In vivo identification of different breast cancer phenotypes can improve our ability to detect more aggressive regions within the tumor and evaluate treatment response

1) To list shortcomings of mammographic breast cancer screening. 2) To describe methods of non-mammographic breast cancer screening. 3) To list possible advantages and disadvantages of non-mammographic breast cancer screening.

Critics of breast MRI point to the high cost of the exam, the false-positive rates and the detection of indolent breast cancers. A shorter MRI may be cheaper and still allow the detection of breast cancer. The purpose of our study was to evaluate the ability of an MRI protocol with one post-contrast (and subtracted) sequence at 90 seconds to detect biologically significant cancers.

An IRB approved retrospective review of 103 women with 180 findings who underwent a breast MRI at 3T was performed by 2 readers. 90 women were newly diagnosed with breast cancer and 13 were asymptomatic high-risk women. Prior to this study, each reader interpreted 228 abridged MRI exams. The scan time for the 3 T1-scans was 4 minutes; the scan time for the T2-sequence was 4 minutes. Final BIRADS assessment and confidence score was assessed for each lesion. Comparison was made to the original diagnostic interpretation.

Of 125 cancers, 4 were foci, 86 were masses, 25 were NME and 10 were categorized in the original report as both masses and NME. Seventy-nine were IDC, 10 were ILC, 23 were DCIS, and 13 were IDC and DCIS. The mean size was 1.7 cm (range 0.4 - 8.6 cm). All 11 mammographically occult contralateral malignancies were detected. Cancers with rim enhancement, spiculated margins or washout kinetics were identified with high confidence by both readers. The sensitivity for reader 1 was 99.2% (CI 95.0-99.9%) and reader 2 was 96% (CI 90.4-98.5%). Of 6 missed cancers, one was IDC, one was ILC and 4 were DCIS. Both invasive cancers were moderately differentiated and all DCIS were intermediate grade. Their mean size was 1.1 cm, range 0.4 - 2cm. All 6 malignancies were seen on the 2nd post-contrast scan and they had Type 1 kinetics. Three were NME and 3 were masses. Eight of 55 (14.5%) MR biopsy proven benign lesions were not identified by each reader, although 4 additional findings were identified by both readers.

An abridged breast MRI protocol yielded 98% sensitivity for invasive cancers, 83% sensitivity for DCIS and increased specificity as compared with a routine breast MR exam. Total acquisition time is 7 minutes compared to 35 minutes for the conventional exam.

Almost all biological significant cancers are detected with an abridged MRI protocol.

To review patients undergoing high risk breast MRI due to personal history of pre-menopausal breast cancer and to determine the incidence of additional cancers found.

With Institutional Review Board approval and waiver of informed consent, a retrospective review was conducted to determine patients diagnosed with pre-menopausal breast cancer undergoing screening high risk MRI. 296 High risk MRI exams were performed in 127 patients from 2003 to 2014. Data recorded included patient age and breast density, lesion size on MRI (if applicable), type of biopsy procedure (if applicable), and pathology results (if applicable).

Total number of MRI exams performed per patient ranged from 1 to 8. Average patient age at the time of first cancer diagnosis was 40.6 years (range 19-48). 76% of patients had heterogeneously dense or extremely dense breast tissue. Of 296 exams, there were 68 (23%) suspicious MRI findings. 47 needle biopsy procedures were performed (69%); 3 did not have a biopsy and proceeded to surgery. MRI biopsy was performed in 30 and ultrasound biopsy in 13. One stereotactic biopsy was performed and FNAC was performed in 3. The remaining 18 findings were determined to be benign by targeted ultrasound, were no longer visualized when MRI biopsy was attempted or were followed and remained stable. Pathology revealed 15 malignancies (10 invasive and 5 non-invasive), 30 benign findings, and 5 atypical findings. Of those diagnosed with cancer, average time between diagnoses was 6.6 years (range 3-16). Seven cancers were in the ipsilateral breast; 5 were of the same pathology as the original malignancy and 2 were different. Eight new cancers were diagnosed in the contralateral breast.

Screening MRI in patients with a personal history of pre-menopausal breast cancer detected a new suspicious finding in 23% of exams. Of 127 patients screened, malignancy was detected in 15 (12%). The cancer diagnoses were detected up to 16 years after initial diagnosis, which demonstrates the importance of monitoring these patients.

Monitoring women with a history of pre-menopausal breast cancer is important as these patients are at increased risk of a second breast cancer diagnosis.

To show the diagnostic value of bi-exponential ADC based on IVIM with multiple b-value DWI in benign and malignant breast lesions, in comparison to the conventional mono-exponential apparent diffusion coefficient (ADC) with single b-factor DWI.

32 patients diagnosed with 20 malignant and 15 benign breast lesions were enrolled in the study. Consent form has been obtained prior to the study. Patients underwent DWI at 3.0T with single b-factor range (b=0, 1000 s/mm²) and multiple b-factor range (b=0, 25, 50, 75, 100, 300, 500, 800, 1000, 1200, 2000, 3000 s/mm²). 32 contralateral normal healthy glandular tissues from the same cohort were considered as control. ADC (b=0 and 1000 s/mm²) and IVIM parameters (tissue diffusivity D, pseudo-diffusion coefficient D*, perfusion fraction f) were calculated respectively based on mono-exponential and bi-exponential analysis. The data were compared in between malignant, benign lesions and normal healthy glandular tissues. The diagnostic efficiency of these parameters was evaluated by ROC curve and area under the ROC curve (AUC).

It was found out that ADC and D values of malignant group were significant lower than those of benign group and control group (P<0.05). There were significant differences between benign and malignant group, benign and control group, malignant and control group in value of f (P<0.01, P<0.05, P<0.01 respectively). Importantly, the f value of malignant tumors was significantly higher than that of benign lesions especially in the b-factor range of 0-800 s/mm2 (P = 0.000).D* value was not significantly different between benign and malignant group in the b-factor range of 800-3000 s/mm2. The diagnostic sensitivity of D, ADC, f, D* decreased progressively. D and ADC values showed higher specificity than f, D*. The combination of D and f represented higher sensitivity and specificity.

Quantitative IVIM parameters provide separate information of fast and slow diffusion component by bi-exponential decay model. They can be used in differential diagnosis of benign and malignant lesions.

Multi-b-value DWI has been most simply performed, and IVIM can separately estimate tissue perfusion and diffusivity. Although some questions are remained to be clarified, multi-b-value DWI and IVIM will certainly be of great help for the diagnosis of breast lesions.

To evaluate 3-D fused gadolinium-enhanced and diffusion-weighted images in preoperative assessment of multicentricity, multifocality, and bilaterality in patients with breast carcinoma

72 patients with biopsy-proven breast carcinoma have been sent to MR imaging for preoperative assessment of multicentricity, multifocality, and bilaterality. Based on mastectomy/ lumpectomy specimens, 47 patients had single lesions. 12 multicentric , 10 multifocal, and 3 bilateral carcinoma with a total number of 116 proven malignant foci. The preoperative MR images were post-processed on an advanced workstation to obtain 3D fused images of the unenhanced breast parenchyma, early gadolinium enhancement (coded red), and diffusion-weighted images with b=1500 s/mm2(coded green). To eliminate the T2-shine-through effect, lesions with ADC ≥ 1× 10-3 mm 2 /s were eliminated. The post-processed images were reviewed by an experienced radiologist, blinded to the histopathology, who noted all the lesions with a diameter ≥ 5 mm classifying them into three groups: matched enhancement and diffusion restriction, unmatched diffusion restriction, and unmatched enhancement.

313 Lesions with a diameter ≥ 5 mm have been identified. 101 lesions showed matched enhancement and difusion restriction. Taking matched lesions as indicative of malignancy, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for diagnosis of malignant foci have been 84.5, 98.5, 97, 91, and 93.3 % respectively. Three false positive foci of matched E-DR were due to fibroadenomas. 18 false negative foci have been due to foci of DCIS less than 1 cm in diameter. The method correctly identified all cases of bilaterality and multifocality, Three patients with multicenteric malignancy were diagnosed as individual lesion. Three patients with multicentricity were diagnosed as multifocal because of fibroadenomas. Three patients were correctly classified as multicentric with underestimated number of foci.

Fused gadolinium-enhanced and diffusion-weighted MR images of the breast offer a reasonably accurate assessment of bilaterality, multifocality, and multicentricity in patients with breast carcinoma.

The technique may be useful in patients with breast carcinoma suspected to have multiple lesions. The 3D and color coded images are easy to read and optimal to use by the surgeon for surgical planning.