The Frequency and Causes of Not-Detected Breast Malignancy in Dynamic Contrast-Enhanced MRI

Breast MR is the most sensitive imaging modality, but there are cases of malignant tumors that are not detected in MR. This study evaluated the frequency and main causes of malignant breast lesions not detected in dynamic contrast-enhanced (DCE) MR. A total of 1707 cases of preoperative breast MR performed between 2020 and 2021 were included. Three radiologists individually reviewed the DCE MRs and found not-detected malignancy cases in the MRs. The final cases were decided through consensus. For the selected cases, images other than DCE MRIs, such as mammography, ultrasounds, diffusion-weighted MRs, and, if possible, contrast-enhanced chest CTs, were analyzed. In the final sample, 12 cases were not detected in DCE MR, and the frequency was 0.7% (12/1707). Six cases were not detected due to known non-enhancing histologic features. In four cases, tumors were located in the breast periphery and showed no enhancement in MR. In the remaining two cases, malignant lesions were not identified due to underlying marked levels of BPE. The frequency of not-detected malignancy in DCE MR is rare. Knowing the causes of each case and correlating it with other imaging modalities could be helpful in the diagnosis of breast malignancy in DCE MR.

Modeling the diffusion-weighted imaging signal for breast lesions in the b = 200 to 3000 s/mm2 range: quality of fit and classification accuracy for different representations

PURPOSE

To evaluate different non-Gaussian representations for the diffusion-weighted imaging (DWI) signal in the b-value range 200 to 3000 s/mm² in benign and malignant breast lesions.

METHODS

Forty-three patients diagnosed with benign (n = 18) or malignant (n = 25) tumors of the breast underwent DWI (b-values 200, 600, 1200, 1800, 2400, and 3000 s/mm² ). Six different representations were fit to the average signal from regions of interest (ROIs) at different b-value ranges. Quality of fit was assessed by the corrected Akaike information criterion (AICc), and the Friedman test was used for assessing representation ranks. The area under the curve (AUC) of receiver operating characteristic curves were used to evaluate the power of derived parameters to differentiate between malignant and benign lesions. The lesion ROI was divided in central and peripheral parts to assess potential effect of heterogeneity. Sensitivity to noise-floor correction was also evaluated.

RESULTS

The Padé exponent was ranked as the best based on AICc, whereas 3 models (kurtosis, fractional, and biexponential) achieved the highest AUC = 0.99 for lesion differentiation. The monoexponential model at b_max = 600 s/mm² already provides AUC = 0.96, with considerably shorter acquisition time and simpler analysis. Significant differences between central and peripheral parts of lesions were found in malignant lesions. The mono- and biexponential models were most stable against varying degrees of noise-floor correction.

CONCLUSION

Non-Gaussian representations are required for fitting of the DWI curve at high b-values in breast lesions. However, the added clinical value from the high b-value data for differentiation of benign and malignant lesions is not clear.

Support vector machine for breast cancer classification using diffusion-weighted MRI histogram features: Preliminary study

BACKGROUND

Diffusion-weighted MRI (DWI) is currently one of the fastest developing MRI-based techniques in oncology. Histogram properties from model fitting of DWI are useful features for differentiation of lesions, and classification can potentially be improved by machine learning.

PURPOSE

To evaluate classification of malignant and benign tumors and breast cancer subtypes using support vector machine (SVM).

STUDY TYPE

Prospective.

SUBJECTS

Fifty-one patients with benign (n = 23) and malignant (n = 28) breast tumors (26 ER+, whereof six were HER2+).

FIELD STRENGTH/SEQUENCE

Patients were imaged with DW-MRI (3T) using twice refocused spin-echo echo-planar imaging with echo time / repetition time (TR/TE) = 9000/86 msec, 90 × 90 matrix size, 2 × 2 mm in-plane resolution, 2.5 mm slice thickness, and 13 b-values.

ASSESSMENT

Apparent diffusion coefficient (ADC), relative enhanced diffusivity (RED), and the intravoxel incoherent motion (IVIM) parameters diffusivity (D), pseudo-diffusivity (D*), and perfusion fraction (f) were calculated. The histogram properties (median, mean, standard deviation, skewness, kurtosis) were used as features in SVM (10-fold cross-validation) for differentiation of lesions and subtyping.

STATISTICAL TESTS

Accuracies of the SVM classifications were calculated to find the combination of features with highest prediction accuracy. Mann-Whitney tests were performed for univariate comparisons.

RESULTS

For benign versus malignant tumors, univariate analysis found 11 histogram properties to be significant differentiators. Using SVM, the highest accuracy (0.96) was achieved from a single feature (mean of RED), or from three feature combinations of IVIM or ADC. Combining features from all models gave perfect classification. No single feature predicted HER2 status of ER + tumors (univariate or SVM), although high accuracy (0.90) was achieved with SVM combining several features. Importantly, these features had to include higher-order statistics (kurtosis and skewness), indicating the importance to account for heterogeneity.

DATA CONCLUSION

Our findings suggest that SVM, using features from a combination of diffusion models, improves prediction accuracy for differentiation of benign versus malignant breast tumors, and may further assist in subtyping of breast cancer.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2018;47:1205-1216.

Diffusion Weighted MR Imaging of Breast and Correlation of Prognostic Factors in Breast Cancer

BACKGROUND

Through Diffusion Weighted Imaging (DWI), information related to early molecular changes, changes in the permeability of cell membranes, and early morphologic and physiologic changes such as cell swelling can be obtained.

AIMS

We investigated the correlation between the prognostic factors of breast cancer and apparent diffusion coefficient (ADC) in DWI sequences of malignant lesions.

STUDY DESIGN

Retrospective cross-sectional study.

METHODS

Patients who were referred to our clinic between September 2012 and September 2013, who underwent dynamic breast MRI before or after biopsy and whose biopsy results were determined as malignant, were included in our study. Before the dynamic analysis, DWI sequences were taken. ADC relationship with all prognostic factors was investigated. Pearson correlation test was used to compare the numerical data, while Spearman correlation and Fisher exact tests were used to compare the categorical data. The advanced relationships were evaluated with linear regression analysis and univariate analysis. The efficiency of the parameters was evaluated using ROC analysis. The significance level (P) was accepted as 0.05.

RESULTS

In total, 41 female patients with an average age of 49.4 years (age interval 21-77) and 44 lesions were included into the study. In the Pearson correlation test, no statistically significant difference was determined between ADC and the patient's age and tumor size. In the Spearman correlation test, a statistically significant difference was determined between nuclear grade (NG) and ADC (r=-0.424, p=0.04); no statistically significant correlation was observed between the other prognostic factors with each other and ADC values. In the linear regression analysis, the relationship of NG with ADC was found to be more significant alone than when comparing all parameters (corrected r2=0.196, p=0.005). Further evaluations between the NG and ADC correlation were carried out with ROC analysis. A statistically significant difference was determined when NG 1 separately was compared with NG 2 and 3 (p=0.03). A statistically significant difference was also determined (p=0.05) in the comparison of NG 1 with only NG 3. No statistically significant difference was determined when NG 2 separately was compared with NG 1 and NG 3 and when NG 3 separately was compared with NG 1 and 2 (p=0.431, p=0.097).

CONCLUSION

We found that ADC values obtained by breast DWI showed a higher correlation with the NG of breast cancer, which is an important factor in the patient's treatment. Predictions can be made about NG by analyzing the ADC values. Additional studies are needed, however, and the ADC value of the lesion can be used as a prognostic factor proving the aggressiveness.

Radiofrequency configuration to facilitate bilateral breast (31) P MR spectroscopic imaging and high-resolution MRI at 7 Tesla

PURPOSE

High-resolution MRI combined with phospholipid detection may improve breast cancer grading. Currently, configurations are optimized for either high-resolution imaging or (31) P spectroscopy. To be able to perform both imaging as well as spectroscopy in a single session, we integrated a (1) H receiver array into a (1) H-(31) P transceiver at 7T. To ensure negligible signal loss due to coupling between elements, we investigated the use of a floating decoupling loop to enable bilateral MRI and (31) P MRS.

METHODS

Two quadrature double-tuned radiofrequency coils were designed for bilateral breast MR with active detuning at the (1) H frequency. The two coils were placed adjacent to each other and decoupled for both frequencies with a single resonant floating loop. Sensitivity of the bilateral configuration, facilitating space for a 26-element (1) H receive array, was compared with a transceiver configuration.

RESULTS

The floating loop was able to decouple the elements over 20 dB for both frequencies. Enlargement of the elements, to provide space for the receivers, and the addition of detuning electronics altered the (31) P sensitivity by 0.4 dB.

CONCLUSION

Dynamic contrast-enhanced scans of 0.7 mm isotropic, diffusion-weighted imaging, and (31) P MR spectroscopic imaging can be acquired at 7T in a single session as demonstrated in a patient with invasive ductal carcinoma.