Breast Region Segmentation being Convolutional Neural Network in Dynamic Contrast Enhanced MRI.

To propose and evaluate a fully automated technique for quantification of fibroglandular tissue (FGT) and background parenchymal enhancement (BPE) in breast MRI. We propose a fully automated method, where after preprocessing, FGT is segmented in T1-weighted, nonfat-saturated MRI. Incorporating an anatomy-driven prior probability for FGT and robust texture descriptors against intensity variations, our method effectively addresses major image processing challenges, including wide variations in breast anatomy and FGT appearance among individuals. Our framework then propagates this segmentation to dynamic contrast-enhanced (DCE)-MRI to quantify BPE within the segmented FGT regions. Axial and sagittal image data from 40 cancer-unaffected women were used to evaluate our proposed method vs a manually annotated reference standard. High spatial correspondence was observed between the automatic and manual FGT segmentation (mean Dice similarity coefficient 81.14%). The FGT and BPE quantifications (denoted FGT% and BPE%) indicated high correlation (Pearson's r=0.99 for both) between automatic and manual segmentations. Furthermore, the differences between the FGT% and BPE% quantified using automatic and manual segmentations were low (mean differences: -0.66±2.91% for FGT% and -0.17±1.03% for BPE%). When correlated with qualitative clinical BI-RADS ratings, the correlation coefficient for FGT% was still high (Spearman's ρ=0.92), whereas that for BPE was lower (ρ=0.65). Our proposed approach also performed significantly better than a previously validated method for sagittal breast MRI. Our method demonstrated accurate fully automated quantification of FGT and BPE in both sagittal and axial breast MRI. Our results also suggested the complexity of BPE assessment, demonstrating relatively low correlation between segmentation and clinical rating.

Introduction: The amount and degree of the fibroglandular tissue (FGT) enhancement on breast MRI, known as background parenchymal enhancement (BPE), is an emerging marker of breast cancer risk. Higher qualitative and quantitative BPE levels have been shown to correlate with both future risk of initial breast cancer diagnosis and breast cancer recurrence after treatment. Furthermore, adjuvant and preventative therapies such as radiation and endocrine therapy have been shown to lower BPE in many women, potentially serving as a marker of treatment efficacy. However, the biological basis of BPE remains unclear and elucidation of this mechanism could identify novel targeted prevention strategies. The goal of this study was to assess the association of BPE levels with markers of inflammation, vascularity, proliferation, and estrogen sensitivity in a cohort of high-risk women. Materials and Methods: In this IRB-approved retrospective study, we identified asymptomatic high-risk women who underwent at least one dynamic contrast-enhanced (DCE) breast MRI within one year of prophylactic bilateral mastectomy. Qualitative BPE assessments were obtained from clinical reports. Quantitative BPE measures were obtained using in-house semi-automated software: whole breast FGT was first segmented on pre-contrast images, percent enhancement (PE) was calculated for each voxel as (S1-S0)/S0 × 100%, where S0 and S1 are pre- and early post-contrast signal intensity, respectively, and quantitative BPE was then defined as the mean PE of the FGT. Additionally, the BPE volume ratio was calculated as the fraction of FGT that had a PE >50%. Histopathology markers of inflammation (COX-2), vascularity (VEGF), proliferation (Ki-67), and estrogen receptor status (ER) were prospectively measured on archived mastectomy specimens of each breast. BPE metrics were correlated with histopathological markers using Spearman’s rank correlation, with and without adjustment by menopausal status of the patient. All tests were based on generalized estimating equations models to account for the non-independence of the two breasts per patient. A p-value < 0.05 was considered significant. Results: We identified 56 women (median age: 39; range: 26-63 years) for this study, of which 39% were post-menopausal (N=22). Distribution of qualitative BPE levels were 41% minimal (N=23), 29% mild (N=16), 18% moderate (N=10), and 12% marked (N=7). Qualitative BPE and quantitative BPE markers were moderately positively correlated (r=0.37-0.54, p< 0.01). Table 1 lists the correlations between BPE metrics and histopathology markers. After adjusting for menopausal status, qualitative BPE was positively associated with ER (r=0.27, p=0.02) and VEGF (r=0.24, p< 0.01; r=0.21, p=0.02 adjusted). Quantitative BPE and BPE volume ratio were both positively correlated with VEGF (p< 0.04) and negatively correlated with COX2 (p< 0.05). No significant associations were observed between BPE and Ki-67. Discussion: Our study suggests that increased BPE on MRI positively correlates with increased vascularity and higher estrogen receptor expression but not inflammation or proliferation within normal breast tissue in high-risk women. These findings, along with established associations between BPE and cancer risk, could be useful to select and assess efficacy of targeted preventative treatments. Table. Spearman rank correlations between BPE metrics and histopathological markers. Citation Format: Pegah Khoshpouri, Anum Kazerouni, Sana Parsian, Daniel Hippe, Olivia Walsh, Lisa Koch, Savannah Partridge, Habib Rahbar. The biological basis of breast MRI background parenchymal enhancement in women with high breast cancer risk [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-07-07.

Purpose Background parenchymal enhancement (BPE), the contrast enhancement of normal fibroglandular tissue on dynamic contrast-enhanced breast MRI, is hormonally dependent and fluctuates with the menstrual cycle, lactation, and pregnancy. In patients with breast cancer, both endocrine therapy and chemotherapy can decrease BPE levels during treatment. Notably, a lack of BPE suppression was associated with inferior response after neoadjuvant chemotherapy in hormone receptor-positive (HR+) cancer. In HR+ breast cancer, BPE may be particularly relevant as estrogen fuels the tumor’s growth. However, there is limited clinical research directly comparing BPE and serum estradiol levels. This study aimed to investigate 1) the correlation between BPE and serum estradiol levels in treatment-naïve premenopausal patients with HR+ breast cancer and 2) the association between BPE suppression and serum estradiol suppression during neoadjuvant endocrine therapy (NET). Methods We retrospectively reviewed data from premenopausal patients enrolled in the Endocrine Optimization Pilot (EOP), a sub-study of the I-SPY 2 TRIAL, which tests new endocrine strategies in patients with stage 2/3, MammaPrint (MP) low-risk (index 0 to 1), HR+/HER2-negative (HER2-) breast cancer. Patients with MP High1 (index -0.57 to 0) tumors were eligible if clinically node negative. This study included 58 EOP patients who had both serial serum estradiol levels (drawn monthly) and BPE data available. These patients started ovarian function suppression (OFS) at various times based on their assigned treatment regimen: before NET start, n=28; at cycle 1, n=16; at cycle 2, n=14. The average estradiol levels over six months on NET were categorized into two groups: suppressed to the postmenopausal range (<10 pg/mL) or non-suppressed (≥10 pg/mL). MRI scans at baseline (before NET) and pre-surgery (after six months of NET) were analyzed. BPE was calculated on the contralateral breast as the mean early (∼150s post-contrast injection) percent enhancement of the central 50% of the axial slices. The Spearman’s correlation coefficient and the Wilcoxon rank-sum test were used to examine the relationship between BPE and estradiol. Results Treatment-naïve estradiol levels were available for 25 of the 58 patients, with the median [1st, 3rd quartile] of 82.7 [32.4, 250] pg/mL. A positive correlation was observed between treatment-naïve estradiol and baseline BPE, with a correlation coefficient of ρ = 0.74 (p < 10-5). There was no significant difference in the treatment-naïve estradiol levels between patients who later showed estradiol suppression to the menopausal range during NET (n=14 [56 %]) and those who did not (n=11 [44 %]) (81.3 pg/mL versus 86.5 pg/mL, respectively (p=0.39)). During NET, estradiol levels were suppressed to the postmenopausal range in 26 of the 58 patients (45%). Estradiol suppression was significantly associated with greater reduction in post-treatment BPE: -42.7% [-54.3, -21.8] BPE reduction for the suppressed estradiol group versus -15.0% [-40.6, 6.85] for the non-suppressed estradiol group (p=0.002). Conclusion This study demonstrates the estrogen dependency of BPE in premenopausal patients with HR+/HER2- breast cancer. Baseline estradiol levels positively correlated with baseline BPE, and the suppression of on-treatment estradiol was associated with greater BPE suppression following NET. Our findings further support exploration of the use of BPE as a reflection of estradiol level under NET and OFS in addition to other MRI biomarkers such as functional tumor volume. These complementary imaging biomarkers could independently reflect both non-malignant and tumor-specific biological processes to improve predicted performance. Citation Format: S. Choi, N. Onishi, P. Metanat, R. Mukhtar, K. Giridhar, M. P. Goetz, L. Huppert, A. Elias, O. Olopade, I-SPY2 Imaging Working Group, I-SPY2 Investigator Network, L. Esserman, J. Chien, N. Hylton. Correlation between MRI background parenchymal enhancement and serum estradiol in premenopausal patients on neoadjuvant endocrine therapy for breast cancer [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS3-07-22.

Magnetic resonance imaging (MRI) plays an important role in breast cancer detection and diagnosis. Breast region segmentation on MRI images is an essential step for many analysis tasks such as the assessment of background parenchymal enhancement (BPE), the analysis of non-mass enhancement (NME) and investigation of the tumor characteristics. However, automated breast segmentation remains a challenge because the noisy background and the blurred boundary between breast and pectoral muscle on MRI images. In this study, we developed a coordinate-guided U-Net (CU-Net) to identify the boundary of the breast on MRI images. The CU-Net applied the breast position information to the segmentation of breast by adding the X coordinate channel and Y coordinate channel to the input image. A dataset of 80 3D bilateral breast MRI scans including 64 cases in training/validation set and 16 cases in test set were collected. All the input images were resized to 256×256 pixels and the intensity are cropped and rescaled according to the percentile of cumulative histogram to normalize the image. To add X and Y coordinate channels, the middle point of the breast-air boundary was detected as the origin of coordinates. The position of each pixel in the coordinates was then encoded into the input of the network. The CU-Net was evaluated with three measures: the dice coefficient (Dice), the intersection over union (IoU) and the hausdorff distance (Hdist). The results of CU-Net were compared to those obtained with U-Net method. In the test set including 16 cases, the mean Dice, the mean IoU, and the mean Hdist of CUNet were 94.1±1.1%, 92.5±1.3%, and 6.17±1.4 mm, while the corresponding measures by U-Net were 91.4±2.4%, 89.7±1.7%, and 6.98±1.5 mm, respectively. The results showed that our CU-Net was able to segment breast on MRI images reliably and was more accurate comparing to the U-Net.

Purpose Standard treatment for ductal carcinoma in situ (DCIS) involves surgical excision and radiation treatment, and often adjuvant endocrine therapy (ET). However, this may constitute overtreatment for some women. The purpose of this neoadjuvant ET study in patients with biopsy-proven DCIS was to evaluate the relationship between the primary outcome of upstaging at surgery (invasive disease) and changes in the pre- and post-ET MRI appearances of both (1) lesion and (2) background parenchymal enhancement (BPE), with the goal to better inform de-escalated therapy. Methods Outcomes of patients with DCIS who were prospectively enrolled in a neoadjuvant ET trial between 2002 and 2009 were retrospectively analyzed with IRB approval. All patients had dynamic contrast-enhanced breast MRI prior to and following an average of 3.1 months of ET (range 1.9 – 5 months), followed by surgical excision within one year of the second MRI. Pathology reports were used to confirm initial biopsy results and to determine the nature and extent of disease postoperatively. Surgeon and radiologist’s reports were used to assess improvement of lesion after ET based on 1) less prominently enhancing pattern or 2) reduction in size of DCIS lesion after treatment. BPE response to ET on imaging was assessed by two radiologists specialized in breast imaging who compared pre- and post-ET MRI’s, blinded to all imaging reports. The change in BPE of the ipsilateral breast with treatment was scored subjectively as increased, decreased, or unchanged. Readers reached a consensus on discordant classifications of the change in BPE. The correlation between imaging change (of lesion and BPE) and incidence of invasive ductal carcinoma (IDC) at surgery was determined using Fisher’s exact test. Results: Of 34 patients evaluated, six (18%) had IDC at surgery. The prevalence of IDC in each group is shown in Table 2. Those with improvement of lesion on MRI had a lower likelihood of IDC at surgery compared to patients with no imaging improvement (Table 2, p= 0.048). Addition of change in BPE to the evaluation of lesion improvement significantly increased the chance of identifying specific patients who were much more likely to have an underlying invasive cancer at surgery (Table 2, p= 0.032). The risk of having IDC at surgery was considerably higher in the group whose lesion was assessed as not improved and who had decreased BPE scores on MRI (Table 2, p= 0.018). Conclusion: In our unique population of patients with mostly ER+ DCIS receiving pre-operative endocrine therapy, the combination of lack of lesion improvement and reduction in BPE score in blinded reads was significantly associated with the finding of invasive disease at surgical excision. This result may reflect the unmasking of a more aggressive lesion in those with underlying IDC and help identify patients with a higher likelihood of the presence of IDC at surgery. Although this study is small, it suggests that risk-stratifying patients with DCIS based on lesion and parenchymal imaging features associated with treatment response may aid in tailoring therapy for DCIS. We are testing this prospectively in an active surveillance cohort. Table 1. Incidence of IDC at surgery depending on image features (n= 34)Improvement of lesion in MRINo Improvement of lesion in MRIP-value (Fisher’s Exact Test)& No Reduction in BPE& Reduction in BPE& No Reduction in BPE& Reduction in BPEIncidence of IDC at surgery (n)0204Incidence of Not IDC at surgery (n)121024%IDC at surgery when assessing lesion only (%)9.1%40%0.048*%IDC at surgery when assessing lesion AND BPE (%)0%16.7%0%50%0.032*%IDC at surgery when assessing lesion AND BPE (%)8.3%50%0.018* Citation Format: Paul Kim, Heather I. Greenwood, Rita I. Freimanis, Gillian L Hirst, Megan Fischer-Colbrie, Jessica Gibbs, Nola M Hylton, Christina Yau, Eun-Sil Shelley Hwang, Laura J. Esserman, Rita A Mukhtar. Improved early stratification of invasive cancer risk using MRI in a ductal carcinoma in-situ short-term active surveillance cohort treated with neoadjuvant endocrine therapy [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-16-06.

Manual segmentation in lumbar paraspinal muscle studies is time-consuming, subject to variability, and limits comprehensive evaluation. To solve the problems, we aim to develop a deep learning algorithm for the automatic segmentation and multi-parameter quantification of eight paraspinal muscles in lumbar CT. We collected CT images covering the range from the superior endplate of the L1 vertebra to the inferior endplate of the S1 vertebra. The dataset was partitioned into training, validation, and test sets in a ratio of 7:1:2. Six convolutional neural networks were employed to automatically segment the bilateral psoas major, quadratus lumborum, erector spinae, and multifidus across all slices. Model performance was evaluated using the Dice Similarity Coefficient (DSC), Hausdorff Distance (HD), and mean Intersection over Union (mIoU). Muscle parameters including cross-sectional area, muscle volume, fat infiltration, CT density, and paraspinal muscle index were calculated. The Intraclass Correlation Coefficient [ICC(3,1)] was used to assess agreement between model's segmentations and manual ground truth. 100 lumbar CT scans (age, 55.02 ± 16.2; 62 female) were used for model development. The mean DSC across all six models ranged from 0.754 to 0.903. TransUNet demonstrated the best overall (DSC, 0.903; 95% CI: 0.902, 0.913) (HD, 9.934mm; 95% CI: 9.815, 11.902) (mIoU, 0.841; 95% CI: 0.839, 0.850). For individual muscles segmented by TransUNet, the highest DSC was 0.936 (95% CI: 0.934-0.938) for the right multifidus. The mean ICC for all parameters was 0.931. The highest ICC value was 0.999 (p < 0.001). The deep learning tool enables automatic and accurate segmentation of eight lumbar paraspinal muscles. Muscle parameter measurements derived from model's segmentations and manual ground truth exhibit high agreement. Accurate segmentation and quantification of paraspinal muscle parameters facilitates large-scale epidemiological studies of paraspinal muscles and spine-related diseases, overcoming the limitations in manual methods.

Amount of fibroglandular tissue (FGT) and background parenchymal enhancement (BPE) in dynamic contrast enhanced magnetic resonance (DCE-MR) images are two important indices for breast cancer risk assessment in the clinical practice. The purpose of this study is to develop and evaluate a fully automated scheme for quantitative analysis of FGT and BPE in DCE-MR images. Our fully automated method consists of three steps, i.e., segmentation of whole breast, fibroglandular tissues, and enhanced fibroglandular tissues. Based on the volume of interest extracted automatically, dynamic programming method was applied in each 2-D slice of a 3-D MR scan to delineate the chest wall and breast skin line for segmenting the whole breast. This step took advantages of the continuity of chest wall and breast skin line across adjacent slices. We then further used fuzzy c-means clustering method with automatic selection of cluster number for segmenting the fibroglandular tissues within the segmented whole breast area. Finally, a statistical method was used to set a threshold based on the estimated noise level for segmenting the enhanced fibroglandular tissues in the subtraction images of pre- and post-contrast MR scans. Based on the segmented whole breast, fibroglandular tissues, and enhanced fibroglandular tissues, FGT and BPE were automatically computed. Preliminary results of technical evaluation and clinical validation showed that our fully automated scheme could obtain good segmentation of the whole breast, fibroglandular tissues, and enhanced fibroglandular tissues to achieve accurate assessment of FGT and BPE for quantitative analysis of breast cancer risk.

Relationship Between Patient Age, Mammographic Density and Background Parenchymal Enhancement on Contrast-Enhanced Mammography.

Breast tissue enhances on contrast MRI and is called background parenchymal enhancement (BPE). Having high BPE has been associated with an increased risk of breast cancer. We examined the relationship between BPE and the amount of fibroglandular tissue on MRI (MRI-FGT) and breast cancer risk factors. This was a cross-sectional study of 415 women without breast cancer undergoing contrast-enhanced breast MRI at Memorial Sloan Kettering Cancer Center. All women completed a questionnaire assessing exposures at the time of MRI. Prevalence ratios (PR) and 95% confidence intervals (CI) describing the relationship between breast cancer risk factors and BPE and MRI-FGT were generated using modified Poisson regression. In multivariable-adjusted models a positive association between body mass index (BMI) and BPE was observed, with a 5-unit increase in BMI associated with a 14% and 44% increase in prevalence of high BPE in pre- and post-menopausal women, respectively. Conversely, a strong inverse relationship between BMI and MRI-FGT was observed in both pre- (PR = 0.66, 95% CI 0.57, 0.76) and post-menopausal (PR = 0.66, 95% CI 0.56, 0.78) women. Use of preventive medication (e.g., tamoxifen) was associated with having low BPE, while no association was observed for MRI-FGT. BPE is an imaging marker available from standard contrast-enhanced MRI, that is influenced by endogenous and exogenous hormonal exposures in both pre- and post-menopausal women.

Stereotactic radiosurgery (SRS) treatment planning requires accurate delineation of brain metastases, a task that can be tedious and time-consuming. Although studies have explored the use of convolutional neural networks (CNNs) in magnetic resonance imaging (MRI) for automatic brain metastases delineation, none of these studies have performed clinical evaluation, raising concerns about clinical applicability. This study aimed to develop an artificial intelligence (AI) tool for the automatic delineation of single brain metastasis that could be integrated into clinical practice. Data from 426 patients with postcontrast T1-weighted MRIs who underwent SRS between March 2007 and August 2019 were retrospectively collected and divided into training, validation, and testing cohorts of 299, 42, and 85 patients, respectively. Two Gamma Knife (GK) surgeons contoured the brain metastases as the ground truth. A novel 2.5D CNN network was developed for single brain metastasis delineation. The mean Dice similarity coefficient (DSC) and average surface distance (ASD) were used to assess the performance of this method. The mean DSC and ASD values were 88.34%±5.00% and 0.35±0.21 mm, respectively, for the contours generated with the AI tool based on the testing set. The DSC measure of the AI tool's performance was dependent on metastatic shape, reinforcement shape, and the existence of peritumoral edema (all P values <0.05). The clinical experts' subjective assessments showed that 415 out of 572 slices (72.6%) in the testing cohort were acceptable for clinical usage without revision. The average time spent editing an AI-generated contour compared with time spent with manual contouring was 74 vs. 196 seconds, respectively (P<0.01). The contours delineated with the AI tool for single brain metastasis were in close agreement with the ground truth. The developed AI tool can effectively reduce contouring time and aid in GK treatment planning of single brain metastasis in clinical practice.

A dual-branch hybrid dilated CNN model for the AI-assisted segmentation of meningiomas in MR images

Breast density (BD) and background parenchymal enhancement (BPE) are important imaging biomarkers for breast cancer (BC) risk. We aim to evaluate longitudinal changes in quantitative BD and BPE in high-risk women undergoing dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), focusing on the effects of age and transition into menopause. A retrospective cohort study analyzed 834 high-risk women undergoing breast DCE-MRI for screening between 2005 and 2020. Quantitative BD and BPE were derived using deep-learning segmentation. Linear mixed-effects models assessed longitudinal changes and the effects of age, menopausal status, weeks since the last menstrual period (LMP-wks), body mass index (BMI), and hormone replacement therapy (HRT) on these imaging biomarkers. BD decreased with age across all menopausal stages, whereas BPE declined with age in postmenopausal women but remained stable in premenopausal women. HRT elevated BPE in postmenopausal women. Perimenopausal women exhibited decreases in both BD and BPE during the menopausal transition, though cross-sectional age at menopause had no significant effect on either measure. Fibroglandular tissue was positively associated with BPE in perimenopausal women. We highlight the dynamic impact of menopause on BD and BPE and correlate well with the known relationship between risk and age at menopause. These findings advance the understanding of imaging biomarkers in high-risk populations and may contribute to the development of improved risk assessment leading to personalized chemoprevention and BC screening recommendations.

Imaging and clinicopathologic characteristics in a contemporary cohort of younger women with newly diagnosed breast cancer

Deep convolutional neural networks (DCNNs) for diagnosis of disease on chest radiographs (CXR) have been shown to be biased against males or females if the datasets used to train them have unbalanced sex representation. Prior work has suggested that DCNNs can predict sex on CXR, which could aid forensic evaluations, but also be a source of bias. To (1) evaluate the performance of DCNNs for predicting sex across different datasets and architectures and (2) evaluate visual biomarkers used by DCNNs to predict sex on CXRs. Chest radiographs were obtained from the Stanford CheXPert and NIH Chest XRay14 datasets which comprised of 224,316 and 112,120 CXRs, respectively. To control for dataset size and class imbalance, random undersampling was used to reduce each dataset to 97,560 images that were balanced for sex. Each dataset was randomly split into training (70%), validation (10%), and test (20%) sets. Four DCNN architectures pre-trained on ImageNet were used for transfer learning. DCNNs were externally validated using a test set from the opposing dataset. Performance was evaluated using area under the receiver operating characteristic curve (AUC). Class activation mapping (CAM) was used to generate heatmaps visualizing the regions contributing to the DCNN's prediction. On the internal test set, DCNNs achieved AUROCs ranging from 0.98 to 0.99. On external validation, the models reached peak cross-dataset performance of 0.94 for the VGG19-Stanford model and 0.95 for the InceptionV3-NIH model. Heatmaps highlighted similar regions of attention between model architectures and datasets, localizing to the mediastinal and upper rib regions, as well as to the lower chest/diaphragmatic regions. DCNNs trained on two large CXR datasets accurately predicted sex on internal and external test data with similar heatmap localizations across DCNN architectures and datasets. These findings support the notion that DCNNs can leverage imaging biomarkers to predict sex and potentially confound the accurate prediction of disease on CXRs and contribute to biased models. On the other hand, these DCNNs can be beneficial to emergency radiologists for forensic evaluations and identifying patient sex for patients whose identities are unknown, such as in acute trauma.

3,3-Diindolylmethane (DIM): a nutritional intervention and its impact on breast density in healthy BRCA carriers. A prospective clinical trial