Breast Lesion Classification with Multiparametric Breast MRI Using Radiomics and Machine Learning: A Comparison with Radiologists' Performance.

This multicenter retrospective study compared the performance of radiomics analysis coupled with machine learning (ML) with that of radiologists for the classification of breast tumors. A total of 93 consecutive women (mean age: 49 ± 12 years) with 104 histopathologically verified enhancing lesions (mean size: 22.8 ± 15.1 mm), classified as suspicious on multiparametric breast MRIs were included. Two experienced breast radiologists assessed all of the lesions, assigning a Breast Imaging Reporting and Database System (BI-RADS) suspicion category, providing a diffusion-weighted imaging (DWI) score based on lesion signal intensity, and determining the apparent diffusion coefficient (ADC). Ten predictive models for breast lesion discrimination were generated using radiomic features extracted from the multiparametric MRI. The area under the receiver operating curve (AUC) and the accuracy were compared using McNemar's test. Multiparametric radiomics with DWI score and BI-RADS (accuracy = 88.5%; AUC = 0.93) and multiparametric radiomics with ADC values and BI-RADS (accuracy= 88.5%; AUC = 0.96) models showed significant improvements in diagnostic accuracy compared to the multiparametric radiomics (DWI + DCE data) model (p = 0.01 and p = 0.02, respectively), but performed similarly compared to the multiparametric assessment by radiologists (accuracy = 85.6%; AUC = 0.03; p = 0.39). In conclusion, radiomics analysis coupled with the ML of multiparametric MRI could assist in breast lesion discrimination, especially for less experienced readers of breast MRIs.

Differentiation Between Benign and Metastatic Breast Lymph Nodes Using Apparent Diffusion Coefficients.

The aim of this study was to determine the range of apparent diffusion coefficient (ADC) values for benign axillary lymph nodes in contrast to malignant axillary lymph nodes, and to define the optimal ADC thresholds for three different ADC parameters (minimum, maximum, and mean ADC) in differentiating between benign and malignant lymph nodes. This retrospective study included consecutive patients who underwent breast MRI from January 2017-December 2020. Two-year follow-up breast imaging or histopathology served as the reference standard for axillary lymph node status. Area under the receiver operating characteristic curve (AUC) values for minimum, maximum, and mean ADC (min ADC, max ADC, and mean ADC) for benign vs malignant axillary lymph nodes were determined using the Wilcoxon rank sum test, and optimal ADC thresholds were determined using Youden's Index. The final study sample consisted of 217 patients (100% female, median age of 52 years (range, 22-81), 110 with benign axillary lymph nodes and 107 with malignant axillary lymph nodes. For benign axillary lymph nodes, ADC values (×10-3 mm2/s) ranged from 0.522-2.712 for mean ADC, 0.774-3.382 for max ADC, and 0.071-2.409 for min ADC; for malignant axillary lymph nodes, ADC values (×10-3 mm2/s) ranged from 0.796-1.080 for mean ADC, 1.168-1.592 for max ADC, and 0.351-0.688 for min ADC for malignant axillary lymph nodes. While there was a statistically difference in all ADC parameters (p<0.001) between benign and malignant axillary lymph nodes, boxplots illustrate overlaps in ADC values, with the least overlap occurring with mean ADC, suggesting that this is the most useful ADC parameter for differentiating between benign and malignant axillary lymph nodes. The mean ADC threshold that resulted in the highest diagnostic accuracy for differentiating between benign and malignant lymph nodes was 1.004×10-3 mm2/s, yielding an accuracy of 75%, sensitivity of 71%, specificity of 79%, positive predictive value of 77%, and negative predictive value of 74%. This mean ADC threshold is lower than the European Society of Breast Imaging (EUSOBI) mean ADC threshold of 1.300×10-3 mm2/s, therefore suggesting that the EUSOBI threshold which was recently recommended for breast tumors should not be extrapolated to evaluate the axillary lymph nodes.

Assessing PD-L1 Expression Status Using Radiomic Features from Contrast-Enhanced Breast MRI in Breast Cancer Patients: Initial Results.

The purpose of this retrospective study was to assess whether radiomics analysis coupled with machine learning (ML) based on standard-of-care dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can predict PD-L1 expression status in patients with triple negative breast cancer, and to compare the performance of this approach with radiologist review. Patients with biopsy-proven triple negative breast cancer who underwent pre-treatment breast MRI and whose PD-L1 status was available were included. Following 3D tumor segmentation and extraction of radiomic features, radiomic features with significant differences between PD-L1+ and PD-L1- patients were determined, and a final predictive model to predict PD-L1 status was developed using a coarse decision tree and five-fold cross-validation. Separately, all lesions were qualitatively assessed by two radiologists independently according to the BI-RADS lexicon. Of 62 women (mean age 47, range 31-81), 27 had PD-L1- tumors and 35 had PD-L1+ tumors. The final radiomics model to predict PD-L1 status utilized three MRI parameters, i.e., variance (FO), run length variance (RLM), and large zone low grey level emphasis (LZLGLE), for a sensitivity of 90.7%, specificity of 85.1%, and diagnostic accuracy of 88.2%. There were no significant associations between qualitative assessed DCE-MRI imaging features and PD-L1 status. Thus, radiomics analysis coupled with ML based on standard-of-care DCE-MRI is a promising approach to derive prognostic and predictive information and to select patients who could benefit from anti-PD-1/PD-L1 treatment.

Radiomics and Machine Learning with Multiparametric Breast MRI for Improved Diagnostic Accuracy in Breast Cancer Diagnosis.

The purpose of this multicenter retrospective study was to evaluate radiomics analysis coupled with machine learning (ML) of dynamic contrast-enhanced (DCE) and diffusion-weighted imaging (DWI) radiomics models separately and combined as multiparametric MRI for improved breast cancer detection. Consecutive patients (Memorial Sloan Kettering Cancer Center, January 2018-March 2020; Medical University Vienna, from January 2011-August 2014) with a suspicious enhancing breast tumor on breast MRI categorized as BI-RADS 4 and who subsequently underwent image-guided biopsy were included. In 93 patients (mean age: 49 years ± 12 years; 100% women), there were 104 lesions (mean size: 22.8 mm; range: 7-99 mm), 46 malignant and 58 benign. Radiomics features were calculated. Subsequently, the five most significant features were fitted into multivariable modeling to produce a robust ML model for discriminating between benign and malignant lesions. A medium Gaussian support vector machine (SVM) model with five-fold cross validation was developed for each modality. A model based on DWI-extracted features achieved an AUC of 0.79 (95% CI: 0.70-0.88), whereas a model based on DCE-extracted features yielded an AUC of 0.83 (95% CI: 0.75-0.91). A multiparametric radiomics model combining DCE- and DWI-extracted features showed the best AUC (0.85; 95% CI: 0.77-0.92) and diagnostic accuracy (81.7%; 95% CI: 73.0-88.6). In conclusion, radiomics analysis coupled with ML of multiparametric MRI allows an improved evaluation of suspicious enhancing breast tumors recommended for biopsy on clinical breast MRI, facilitating accurate breast cancer diagnosis while reducing unnecessary benign breast biopsies.

Multidimensional Diffusion Magnetic Resonance Imaging for Characterization of Tissue Microstructure in Breast Cancer Patients: A Prospective Pilot Study.

Diffusion-weighted imaging is a non-invasive functional imaging modality for breast tumor characterization through apparent diffusion coefficients. Yet, it has so far been unable to intuitively inform on tissue microstructure. In this IRB-approved prospective study, we applied novel multidimensional diffusion (MDD) encoding across 16 patients with suspected breast cancer to evaluate its potential for tissue characterization in the clinical setting. Data acquired via custom MDD sequences was processed using an algorithm estimating non-parametric diffusion tensor distributions. The statistical descriptors of these distributions allow us to quantify tissue composition in terms of metrics informing on cell densities, shapes, and orientations. Additionally, signal fractions from specific cell types, such as elongated cells (bin1), isotropic cells (bin2), and free water (bin3), were teased apart. Histogram analysis in cancers and healthy breast tissue showed that cancers exhibited lower mean values of "size" (1.43 ± 0.54 × 10-3 mm2/s) and higher mean values of "shape" (0.47 ± 0.15) corresponding to bin1, while FGT (fibroglandular breast tissue) presented higher mean values of "size" (2.33 ± 0.22 × 10-3 mm2/s) and lower mean values of "shape" (0.27 ± 0.11) corresponding to bin3 (p < 0.001). Invasive carcinomas showed significant differences in mean signal fractions from bin1 (0.64 ± 0.13 vs. 0.4 ± 0.25) and bin3 (0.18 ± 0.08 vs. 0.42 ± 0.21) compared to ductal carcinomas in situ (DCIS) and invasive carcinomas with associated DCIS (p = 0.03). MDD enabled qualitative and quantitative evaluation of the composition of breast cancers and healthy glands.

Diagnostic value of radiomics and machine learning with dynamic contrast-enhanced magnetic resonance imaging for patients with atypical ductal hyperplasia in predicting malignant upgrade.

PURPOSE: To investigate whether radiomics features extracted from magnetic resonance imaging (MRI) of patients with biopsy-proven atypical ductal hyperplasia (ADH) coupled with machine learning can differentiate high-risk lesions that will upgrade to malignancy at surgery from those that will not, and to determine if qualitatively and semi-quantitatively assessed imaging features, clinical factors, and image-guided biopsy technical factors are associated with upgrade rate. METHODS: This retrospective study included 127 patients with 139 breast lesions yielding ADH at biopsy who were assessed with multiparametric MRI prior to biopsy. Two radiologists assessed all lesions independently and with a third reader in consensus according to the BI-RADS lexicon. Univariate analysis and multivariate modeling were performed to identify significant radiomic features to be included in a machine learning model to discriminate between lesions that upgraded to malignancy on surgery from those that did not. RESULTS: Of 139 lesions, 28 were upgraded to malignancy at surgery, while 111 were not upgraded. Diagnostic accuracy was 53.6%, specificity 79.2%, and sensitivity 15.3% for the model developed from pre-contrast features, and 60.7%, 86%, and 22.8% for the model developed from delta radiomics datasets. No significant associations were found between any radiologist-assessed lesion parameters and upgrade status. There was a significant correlation between the number of specimens sampled during biopsy and upgrade status (p = 0.003). CONCLUSION: Radiomics analysis coupled with machine learning did not predict upgrade status of ADH. The only significant result from this analysis is between the number of specimens sampled during biopsy procedure and upgrade status at surgery.

Using Deep Learning to Improve Nonsystematic Viewing of Breast Cancer on MRI.

OBJECTIVE: To investigate the feasibility of using deep learning to identify tumor-containing axial slices on breast MRI images. METHODS: This IRB-approved retrospective study included consecutive patients with operable invasive breast cancer undergoing pretreatment breast MRI between January 1, 2014, and December 31, 2017. Axial tumor-containing slices from the first postcontrast phase were extracted. Each axial image was subdivided into two subimages: one of the ipsilateral cancer-containing breast and one of the contralateral healthy breast. Cases were randomly divided into training, validation, and testing sets. A convolutional neural network was trained to classify subimages into "cancer" and "no cancer" categories. Accuracy, sensitivity, and specificity of the classification system were determined using pathology as the reference standard. A two-reader study was performed to measure the time savings of the deep learning algorithm using descriptive statistics. RESULTS: Two hundred and seventy-three patients with unilateral breast cancer met study criteria. On the held-out test set, accuracy of the deep learning system for tumor detection was 92.8% (648/706; 95% confidence interval: 89.7%-93.8%). Sensitivity and specificity were 89.5% and 94.3%, respectively. Readers spent 3 to 45 seconds to scroll to the tumor-containing slices without use of the deep learning algorithm. CONCLUSION: In breast MR exams containing breast cancer, deep learning can be used to identify the tumor-containing slices. This technology may be integrated into the picture archiving and communication system to bypass scrolling when viewing stacked images, which can be helpful during nonsystematic image viewing, such as during interdisciplinary tumor board meetings.

Pharmacokinetic Analysis of Dynamic Contrast-Enhanced Magnetic Resonance Imaging at 7T for Breast Cancer Diagnosis and Characterization.

The purpose of this study was to investigate whether ultra-high-field dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) of the breast at 7T using quantitative pharmacokinetic (PK) analysis can differentiate between benign and malignant breast tumors for improved breast cancer diagnosis and to predict molecular subtypes, histologic grade, and proliferation rate in breast cancer. In this prospective study, 37 patients with 43 lesions suspicious on mammography or ultrasound underwent bilateral DCE-MRI of the breast at 7T. PK parameters (KTrans, kep, Ve) were evaluated with two region of interest (ROI) approaches (2D whole-tumor ROI or 2D 10 mm standardized ROI) manually drawn by two readers (senior reader, R1, and R2) independently. Histopathology served as the reference standard. PK parameters differentiated benign and malignant lesions (n = 16, 27, respectively) with good accuracy (AUCs = 0.655-0.762). The addition of quantitative PK analysis to subjective BI-RADS classification improved breast cancer detection from 88.4% to 97.7% for R1 and 86.04% to 97.67% for R2. Different ROI approaches did not influence diagnostic accuracy for both readers. Except for KTrans for whole-tumor ROI for R2, none of the PK parameters were valuable to predict molecular subtypes, histologic grade, or proliferation rate in breast cancer. In conclusion, PK-enhanced BI-RADS is promising for the noninvasive differentiation of benign and malignant breast tumors.

MRI background parenchymal enhancement, fibroglandular tissue, and mammographic breast density in patients with invasive lobular breast cancer on adjuvant endocrine hormonal treatment: associations with survival.

BACKGROUND: To investigate if baseline and/or changes in contralateral background parenchymal enhancement (BPE) and fibroglandular tissue (FGT) measured on magnetic resonance imaging (MRI) and mammographic breast density (MD) can be used as imaging biomarkers for overall and recurrence-free survival in patients with invasive lobular carcinomas (ILCs) undergoing adjuvant endocrine treatment. METHODS: Women who fulfilled the following inclusion criteria were included in this retrospective HIPAA-compliant IRB-approved study: unilateral ILC, pre-treatment breast MRI and/or mammography from 2000 to 2010, adjuvant endocrine treatment, follow-up MRI, and/or mammography 1-2 years after treatment onset. BPE, FGT, and mammographic MD of the contralateral breast were independently graded by four dedicated breast radiologists according to BI-RADS. Associations between the baseline levels and change in levels of BPE, FGT, and MD with overall survival and recurrence-free survival were assessed using Kaplan-Meier survival curves and Cox regression analysis. RESULTS: Two hundred ninety-eight patients (average age = 54.1 years, range = 31-79) fulfilled the inclusion criteria. The average follow-up duration was 11.8 years (range = 2-19). Baseline and change in levels of BPE, FGT, and MD were not significantly associated with recurrence-free or overall survival. Recurrence-free and overall survival were affected by histological subtype (p < 0.0001), number of metastatic axillary lymph nodes (p < 0.0001), age (p = 0.01), and adjuvant endocrine treatment duration (p < 0.001). CONCLUSIONS: Qualitative evaluation of BPE, FGT, and mammographic MD changes cannot predict which patients are more likely to benefit from adjuvant endocrine treatment.

Multimodality imaging of intrauterine devices with an emphasis on the emerging role of 3-dimensional ultrasound.

The intrauterine contraceptive device (IUD) is one of the most widely used reversible contraception methods throughout the world. With advancing technology, it has rapidly gained acceptance through its increased effectiveness and practicality compared with more invasive means such as laparoscopic tubal ligation. This pictorial essay will present the IUDs most commonly used today. It will illustrate both normal and abnormal positions of IUDs across all cross-sectional imaging modalities including 2-dimensional ultrasound, computed tomography, and magnetic resonance imaging, with a focus on the emerging role of 3-dimensional ultrasound as the modality of choice.

Characterization of subsets of CD4+ memory T cells reveals early branched pathways of T cell differentiation in humans.

The pathways for differentiation of human CD4(+) T cells into functionally distinct subsets of memory cells in vivo are unknown. The identification of these subsets and pathways has clear implications for the design of vaccines and immune-targeted therapies. Here, we show that populations of apparently naive CD4(+) T cells express the chemokine receptors CXCR3 or CCR4 and demonstrate patterns of gene expression and functional responses characteristic of memory cells. The proliferation history and T cell receptor repertoire of these chemokine-receptor(+) cells suggest that they are very early memory CD4(+) T cells that have "rested down" before acquiring the phenotypes described for "central" or "effector" memory T cells. In addition, the chemokine-receptor(+) "naive" populations contain Th1 and Th2 cells, respectively, demonstrating that Th1/Th2 differentiation can occur very early in vivo in the absence of markers conventionally associated with memory cells. We localized ligands for CXCR3 and CCR4 to separate foci in T cell zones of tonsil, suggesting that the chemokine-receptor(+) subsets may be recruited and contribute to segregated, polarized microenvironments within lymphoid organs. Importantly, our data suggest that CD4(+) T cells do not differentiate according to a simple schema from naive --> CD45RO(+) noneffector/central memory --> effector/effector memory cells. Rather, developmental pathways branch early on to yield effector/memory populations that are highly heterogeneous and multifunctional and have the potential to become stable resting cells.