| Computer # |
| 4591.
 | 113 | Comparison of two deep learning models for contrast agent dose reduction in dynamic contrast enhanced breast MRI Teresa Lemainque1, Luisa Huck1, Gustav Müller-Franzes2, Maike Bode1, Sven Nebelung1, Christiane Kuhl1, and Daniel Truhn1
1Diagnostic and Interventional Radiology, University Hospital RWTH Aachen, Aachen, Germany, 2Diagnostic and Interventional Radiology, Uniklinik RWTH Aachen University, Aachen, Germany Keywords: Breast, Machine Learning/Artificial Intelligence Motivation: In the context of MRI-based breast cancer screening, reducing contrast agent dose is desirable. However, this yields decreased contrast-to-noise ratio in dynamic-contrast-enhanced subtraction images. Goal(s): This work aimed to compare two deep learning techniques, diffusion probabilistic models (DDPM) and general adversarial networks (GAN), for retrospective contrast enhancement of low-dose breast MRI subtraction images. Approach: Training and testing was performed on virtual low dose subtraction images, which we generated by subjecting original subtraction images to different amounts of noise. Results: Both DDPM and GAN could denoise these images; however, neither model was superior over the other across all tested dose levels and evaluation metrics. Impact: Diffusion
probabilistic models and general adversarial networks can retrospectively
enhance the signal of virtual low-dose images. They may supplement imaging with
reduced doses in the future; yet, further development and validation on real
low dose images are warranted. |
| 4592.
 | 114 | Whole-tumor histogram models based on quantitative maps from SyMRI for predicting axillary lymph node status in invasive ductal breast cancer Fang Zeng1, Zheting Yang1, Xiaoxue Tang1, Lin Lin1, Pu-Yeh Wu2, and Yunjing Xue1
1Department of Radiology, Fujian Medical University Union Hospital, Fuzhou, Fujian, China, 2GE Healthcare, Beijing, China Keywords: Breast, Breast Motivation: Breast cancer is closely associated with ALN status, influencing prognosis. Sentinel lymph node (SLN) biopsy, a common ALN staging method, has limitations. Goal(s): This study aimed to explore a non-invasive predictive approach for ALN status in IDC patients using SyMRI images and histogram analysis. Approach: We included 212 patients, and compared the performance of SyMRI histogram models in differentiating N0 and N+ groups (further divided into N1 and N2-3). Results: Combining quantitative map features with clinical data achieved the highest diagnostic accuracy. Additionally, specific histogram features were found to differ significantly between N1 and N2-3 groups. Conventional parameters were less discriminative. Impact: We demonstrated efficacy of histogram analysis of SyMRI as a non-invasive method for predicting ALN status. Model combining SyMRI quantitative maps and clinical features yielded satisfactory performance, highlighting the potential of our proposed model in ALN management. |
| 4593.
 | 115 | Precision Diagnosis of BI-RADS4 Breast Lesions: A Promising Approach with DCE and 3D-MIP Parameters hongbing liang1, lina zhang1, ning ning1, siqi zhao1, yuanfei li1, yueqi wu1, qingwei song1, haonan guan2, and lizhi xie2
1First Affiliated Hospital of Dalian Medical University, Dalian, CHINA, Dalian, China, 2GE Healthcare, MR Research China, Beijing, China, Beijing, China Keywords: Breast, Breast Motivation: Breast cancer is a significant health concern for women. Accurate diagnosis of BI-RADS4 lesions is challenging, necessitating improved diagnostic indicators. Goal(s): This study aims to enhance BI-RADS4 breast lesion diagnosis using DCE and 3D-MIP parameters, providing more precise insights. Approach: We analyzed various parameters, identified independent factors, and combined vascular diameter difference with Slopemax for optimal diagnosis. Results: The combination demonstrated superior diagnostic efficiency, differentiating benign from malignant lesions effectively. Impact: This approach holds promise for early breast cancer diagnosis and improved patient care, offering clinicians a valuable tool for enhanced precision in BI-RADS4 lesion evaluation. |
| 4594.
 | 116 | MRI Diagnosis of Lesions Presenting as Architectural Distortion on DBT: Comparison of Diagnostic Performance Using BI-RADS and Radiomics Models Jiejie Zhou1,2, Xiao Chen1, Yang Zhang2, Yan-lin Liu2, Yong Pan1, Jeon-Hor Chen2, Guoquan Cao1, Meihao Wang1, and Min-ying Su3
1First affiliated hospital of Wenzhou Medical University, Wenzhou, China, 2University of California, Irvine, Irvine, CA, United States, 3University of California, Irvine, Irvine, China Keywords: Breast, Breast Motivation: Diagnosis of lesions shown as architectural distortion (AD) on DBT is challenging, and breast MRI may help. Goal(s): To compare the diagnostic performance of 60 cases using reading based on BI-RADS of DBT and MRI, Kaiser score, and radiomics models. Approach: In addition to comparing the diagnostic performance, features shown on DBT and MRI, and the distribution in different MRI BI-RADS categories, were reported. Results: The malignant rate of AD varied in associated features and MRI-RADS groups. MRI showed better diagnostic performance than DBT. When using radiomics models, the accuracy was almost the same, but the AUC of DBT+MRI fused model improved. Impact: Diagnosis of lesions presenting as AD on DBT can
be improved with more understanding of associated features, as well as the predictive
features based on the supplementary MRI. |
| 4595.
 | 117 | Development and external validation of a combined clinical-mammographic-MRI model for differentiating benign and malignant NME breast lesions Linhua Wu1, Wei Yang1, and Jian Li1
1General Hospital of Ningxia Medical University, Yinchuan, China Keywords: Breast, Breast, Clinical, Mammography, MRI, Nonmass enhancement breast lesions Motivation: Differential diagnosis of nonmass enhancement (NME) breast lesions is difficult. Goal(s): A combined clinical-mammographic-MRI based on DWI model can distinguish benign and malignant NME lesions. Approach: We retrospectively enrolled consecutive female patients with NME breast lesions who underwent pretreatment MG and breast MRI as the development cohort and prospectively collected eligible candidates as an internal validation group and an external validation group at our centre. A combined discriminatory model was developed through multivariable logistic regression and was validated internally and externally. Results: The combined model incorporating mammography, MRI, and clinical variables showed good discriminability in the development, internal validation, and external validation cohorts. Impact: Based on SHAP analysis, suspicious calcification on mammography and internal NME patterns were significant contributors to the performance of the model, and radiologists and clinicians should improve their awareness of complementary mammography to improve diagnostic performance in classifying NME lesions. |
| 4596.
 | 118 | Reliability and repeatability of texture features extracted from quantitative T1 & T2 of fresh breast tumour specimens at 3T Kangwa Alex Nkonde1,2, Sai Man Cheung2, Nicholas Senn2, Ehab Husain3, Yazan Masannat4, and Jiabao He1,2
1Newcastle Magnetic Resonance Centre, Translational and Clinical Research Institute, Faculty of Medical Sciences, Newcastle University, Newcastle, United Kingdom, 2Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, United Kingdom, 3Pathology Department, Aberdeen Royal Infirmary, Aberdeen, United Kingdom, 4Breast Unit, Broomfield Hospital, Chelmsford, United Kingdom Keywords: Breast, Breast Motivation: Quantitative T1/T2 is known to alter in the presence of breast tumours, and texture analysis offers a measure to characterise unique tumour morphology. Goal(s): We aimed to determine the reliability and repeatability of texture features extracted from T1/T2 images across acquisitions. Approach: Five repeated acquisitions of T1/T2 were performed on 20 breast tumours to derive texture features of Mean, Standard deviation, Kurtosis, Skewness, and Entropy. Results: There was excellent reliability and repeatability in all T1 texture features, except moderate reliability in Entropy. There was good to excellent reliability and excellent repeatability for most T2 textures, except Kurtosis and Skewness. Impact: The reliability and repeatability of texture features extracted from relaxation property maps serves as a corner stone towards higher order analysis for breast cancer, to support clinical decision with confidence. |
| 4597.
 | 119 | Microcalcification Detection and Differentiation in Breast Cancer using Ultrashort Echo Time (UTE) MRI Yazan Ayoub1, Sai Man Cheung1, Boddor Maglan1, Nicholas Senn1, Ehab Husain2, Yazan Masannat3, and Jiabao He1,4
1Aberdeen Biomedical Imaging Centre, Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, United Kingdom, 2Pathology Department, Aberdeen Royal Infirmary, Aberdeen, United Kingdom, 3Breast Unit, Broomfield Hospital, Mid and South Essex NHS Trust, Essex, United Kingdom, 4Newcastle Magnetic Resonance Centre, Translational and Clinical Research Institute, Faculty of Medical Sciences, Newcastle University, Newcastle, United Kingdom Keywords: Breast, Breast, Microcalcification UTE Motivation: Ultra short echo time (UTE) has been developed to capture the rapid signal decay of short T2* species, to overcome the insensitivity of conventional MRI towards micro-calcification, a central prognostic marker of breast cancer. Goal(s): To examine the degree of calcification in breast tumour specimens using UTE. Approach: The degree of calcification of whole tumour from 20 specimens freshly excised from female patients with breast cancer was derived using dual-echo UTE protocol, with correlation against histological findings. Results: The degree of calcification was significantly different between malignant and non-calcified tissue, with no significant correlation to Ki-67 and NPI scores. Impact: Ultra-short echo time (UTE) enhances the sensitivity to micro-calcifications in the breast, through dual-echo approach to maximise the signal of short T2* species. Clinical implementation of UTE can enhance assessment of microcalcification and improve prognostic value of breast cancer imaging. |
| 4598.
 | 120 | Preoperative Prediction of Recurrence Risk in Breast Cancer Patients Based on MRI Features Jiejie Zhou1,2, Yang Zhang2, Jinhao Wang3, Yezhi Lin4, Hailing Wang3, Yan-lin Liu2, Jeon-Hor Chen2, Meihao Wang1, and Min-ying Su2
1First affiliated hospital of Wenzhou Medical University, Wenzhou, China, 2University of California, Irvine, Irvine, CA, United States, 3Guangxi Normal University, Guilin, China, 4Wenzhou Medical University, Wenzhou, China Keywords: Breast, Breast Motivation: Early prediction of recurrence risk is essential to treatment decision-making for breast cancer patients. Goal(s): To explore potential predictors of recurrence risk based on MRI features and to construct a preoperatively predictive model of risk. Approach: MRI features of 588 patients were investigated, 397 in training and 191 in testing data. Four machine learning methods were used to construct the predictive model. Results: Multiple lesions, irregular shape, spiculated margin, and peritumor edema were identified as predictive factors and used to construct the model. SVM showed the best predictive performance with AUC 0.87 (95%CI 0.83-0.91) and 0.73 (95%CI 0.75-0.81) in training and testing data. Impact: A preoperative predictive model based on MRI features could be a
valuable tool for predicting recurrence risk and assisting in the personalized
treatment of breast cancer patients. |
| 4599.
 | 121 | Evaluation Molecular Receptors Status in Breast Cancer Using an mpMRI-based Feature Fusion Radiomics Model: Mimicking Radiologists’ Diagnosis Fangrong Liang1, Wanli Zhang1, Jiamin Li1, Xin Zhen2, Xinqing Jiang1, Ruimeng Yang1, and Yongzhou Xu3
1Department of Radiology, The Second Affiliated Hospital, School of Medicine, South China University of Technology, Guangzhou, China, 2School of Biomedical Engineering, Southern Medical University, Guangzhou, China, 3Philips Healthcare, Guangzhou, China Keywords: Breast, Radiomics, Breast Cancer, Molecular Receptors Status, Feature Fusion Radiomics Model Motivation: Investigating the performance of a feature fusion radiomics (RFF) strategy that mimicked the routine diagnostic practices of radiologists in distinguishing different statuses of molecular receptors in breast cancer (BC) preoperatively. Goal(s): Develop an RFF model that incorporates optimal mpMRIs for BC’s molecular receptor status identification. Approach: Constructed and analyzed 150 models to determine the top four optimum sequences for identifying distinct BC’s molecular receptor statuses. Then the optimal single sequence models (Rss) and combined sequences models (RFF) were developed and compared. Results: The RFF model integrating mpMRI radiomics features exhibited promising ability to imitate radiologists’ diagnosis for preoperative identification of BC’s molecular receptors.
Impact: A multiparametric MR-based RFF model, mimicking the radiologists’ daily diagnostic approach, which fused radiomics features with dominant MR sequences, was able to distinguish different molecular receptor statuses of breast cancer. |
| 4600.
 | 122 | Machine Learning with Multiparametric MRI for preoperative prediction of intraductal component in invasive breast cancer Lingsong Meng1, Xin Zhao1, Jinxia Guo2, Lin Lu1, Meiying Cheng1, Qingna Xing1, Honglei Shang1, Penghua Zhang1, Yanyong Shen1, and Xiaoan Zhang1
1The Third Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 2GE Healthcare MR Research, Beijing, China, Beijing, China Keywords: Breast, Breast, Machine Learning Motivation: To predict the presence of an intraductal component (ductal carcinoma in situ, DCIS) in invasive breast cancer (IBC-IC). Goal(s): To improve the preoperative prediction of IBC-IC. Approach: This study was to develop and validate a machine-learning algorithm to preoperatively predict IBC-IC using the multiparametric MRI features. Results: The machine learning model with multiparametric MRI features could provide the individualized probability of IBC-IC and might help to optimize surgical planning for patients with breast cancer before BCS. Impact: This
study developed a prediction model combining a machine-learning algorithm with multiparametric
MRI features to preoperatively predict intraductal component in invasive breast
cancer, which may be beneficial to the preoperative planning of breast-conserving
surgery for early-stage invasive breast cancer. |
| 4601.
 | 123 | Quantitative Background Parenchymal Enhancement: Association with Lifetime Risk Factors on Breast Cancer Screening MRI Ran Yan1,2, Wakana Murakami1, Shabnam Mortazavi1, Tiffany Yu1, Stephanie Lee-Felker1, and Kyunghyun Sung1,2
1Department of Radiological Sciences, University of California, Los Angeles, Los Angeles, CA, United States, 2Department of Bioengineering, University of California, Los Angeles, Los Angeles, CA, United States Keywords: Breast, Screening, Background parenchymal enhancement; Quantitative BPE; Breast cancer; Lifetime risk; BRCA germline mutation Motivation: Breast MRI background parenchymal enhancement (BPE) has been associated with breast cancer risk. The association of BPE with various breast cancer lifetime risk factors remains controversial. Goal(s): Our goal was to determine which risk factors are associated with BPE, specifically quantitative BPE due to variability in qualitative BPE. Approach: We used linear regression to evaluate the association between quantitative BPE and BRCA gene mutation status, age, body mass index (BMI), menopausal status, hormonal treatment, and fibroglandular tissue (FGT) level. Results: Both univariate and multivariate analyses of quantitative BPE showed significant correlations with age, BMI, menopausal status, and FGT level. Impact: Quantitative background parenchymal enhancement (BPE) is correlated with lifetime risk factors, such as age, BMI, menopausal status, and FGT level, on breast cancer screening MRI. This can provide potential insight into the cancer pathophysiological mechanisms underlying lifetime risk models. |
| 4602.
 | 124 | Prediction of axillary lymph nodes metastases in patients with breast cancer : Can synthetic MRI provide additional value to DWI? Xiao Yang1, Zongqiong Sun1, and Weiqiang Dou2
1Affiliated hospital of Jiangnan university, wu xi, China, 2GE Healthcare, Beijing, China Keywords: Breast, Breast, Axillary lymph node; metastasis Motivation: The aim of this study was to investigate whether synthetic MRI-derived quantitative maps can predict metastasis of axillary lymph nodes (ALNs) in breast cancer. Goal(s): Searching for a noninvasive method to predict lymph node metastasis in breast cancer. Approach: 61 breast cancer patients were recruited for the study and the status of ALNs was confirmed by pathology. T1, T2, and PD maps were obtained for each patient. Results: Statistically significant differences in PD, minimal ADC , anatomic features of tumor shape and size were observed between patients with and without metastatic ALNs metastasis. It was shown that PD+Min ADC+shape+size with highest AUC of 0.86. Impact: it can be concluded that syMRI derived relaxation maps may be useful for predicting ALNs status in breast cancer. |
| 4603.
 | 125 | Radiomics of voxelwise DCE-MRI TIC profiles map enables quantifying temporal and spatial hemodynamic heterogeneity in breast lesions Zhou Liu1, Meng Sun2, Bingyu Yao2, Meng Wang1, Ya Ren1, Qian Yang1, Wei Cui3, Na Zhang2, and Dehong Luo1
1National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen, China, 2Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 3MR Research, GE Healthcare, Beijing, China, Beijing, China Keywords: Breast, Cancer, Breast cancer; DCE-MRI; Hemodynamic heterogeneity Motivation: An effective approach that enables simultaneous quantification of spatial and temporal heterogeneity based on DCE-MRI is lacking. Goal(s): To develop a data-driven model-free approach to quantify spatial and temporal hemodynamic heterogeneity. Approach: We introduced radiomics analysis based on voxelwise mapping of DCE-MRI time-intensity-curve (TIC) profiles to quantify temporal and spatial hemodynamic heterogeneity and investigated its value in differentiating malignant and benign breast lesions. Results: Radiomics features and composition ratio of voxelwise TIC profiles showed good performance in differentiating malignant and benign breast lesions. Impact: We provide a novel data-driven model-free approach for visualizing and quantifying temporal and spatial hemodynamic heterogeneity simultaneously, which shows potential for a variety of clinical implications concerning individualized management of breast lesions. |
| 4604.
 | 126 | DCE-MRI Tumor Volumetric Changes Predict Response to Neoadjuvant Immunochemotherapy in Triple Negative Breast Cancer Patients Gaiane Margishvili Rauch1, Tanya Moseley2, Mary Guirguis2, Gary Whitman2, Rosalind Candelaria2, Jessica Leung2, Miral Patel2, Jia Sun3, Huong Le-Petross2, Deanna Lane2, Marion Scoggins4, Frances Perez2, Rania M Mohamed5, Zhan Xu6, Sanaz Pashapoor7, Jong Bum Son6, Ken-Pin Hwang8, Huiqin Chen3, Peng Wei3, Debu Tripathy9, Wei Yang2, Clinton Yam9, Jingfei Ma8, and Beatriz Adrada2
1Abdominal and Breast Imaging, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 2Breast Imaging, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 3Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 4Radiology - Breast Imaging, UT Southwestern Medical Center, Dallas, TX, United States, 5Cancer Systems Imaging, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 6Imaging Physics - Research, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 7Breast Imaging - Research, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 8Imaging Physics, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 9Breast Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, United States Keywords: Breast, Breast, Treatment response prediction, DCE-MRI, biomarkers Motivation: There is unmet need for noninvasive biomarkers for neoadjuvant immunochemotherapy (NICT) response prediction in triple negative breast cancer (TNBC) to guide least toxic and most effective treatment regimens. Goal(s): To evaluate if DCE-MRI tumor volume changes measured early during NICT can predict treatment response. Approach: DCE-MRI tumor volume reduction (TVR) was calculated in 64 TNBC at baseline, after 2 and 4 cycles of NICT and correlated with surgical pathology using ROC analysis. Results: DCE-MRI TVR after 2 cycles of NICT was able to predict pCR with AUC of 0.71 (95%CI:0.57-0.84) and after 4 cycles with AUC of 0.81 (95%CI:0.69-0.92). Impact: DCE-MRI tumor volume changes early during neoadjuvant immunochemotherapy can identify triple negative breast cancer patients with high/low likelihood of pathologic complete response, triaging them to appropriate management for de-escalation trials versus targeted therapies, avoiding unnecessary toxicity of ineffective treatment. |
| 4605.
 | 127 | Enhancing Breast Lesion Diagnosis Through DISCO and Deep Learning Reconstruction-Based DWI Wanjun Xia1, Yong Zhang1, Kaiyu Wang2, Tianyong Xu2, Ruilin Fan1, and Jingliang Cheng1
1Department of Magnetic Resonance, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 2MR Research China, GE Healthcare, Beijing, China Keywords: Breast, Breast, deep learning; DWI: differential diagnosis; DISCO Motivation: With breast cancer now ranking as the predominant global cancer, there is a pressing need to enhance diagnostic accuracy and reduce unnecessary biopsies through the utilization of advanced imaging techniques. Goal(s): Our aim is to augment the precision of breast disease diagnosis by improving the contrast-enhanced MRI and DWI in routine scans. Approach: We developed a model that combines DISCO with deep learning-reconstructed DWI at a b-value of 800 s/mm² for differential diagnosis. Results: The integration of deep learning-reconstructed DWI and DISCO serves to significantly enhance the capability to differentiate between benign and malignant breast conditions. Impact: This advancement directly heightens the diagnostic efficiency of breast cancer within routine scanning sequences, contributing to more effective clinical solutions, and ultimately elevating both the quality of life and survival rates for patients. |
| 4606.
 | 128 | The Application of High Temporal Resolution Semi-quantitative Dynamic Contrast Enhanced MRI in Predicting Ki-67 Expression in Breast Cancer Wen Feng1, Junqiang Lei1, Yuhui Xiong2, Yicong Niu3, Zhifan Li1, Qinqin Ma1, and Zihan Wang4
1Radiology, The First Hospital of Lanzhou University, Lanzhou, China, 2GE HealthCare MR Research, Beijing, China, 3Breast Disease, The First Hospital of Lanzhou University, Lanzhou, China, 4Pathology, The First Hospital of Lanzhou University, Lanzhou, China Keywords: Breast, Breast, MRI; DISCO; DCE; Ki-67 Motivation: It was hoped that a method can come under observation to capture the semi-quantitative hemodynamic characteristics of tumor enhancement, so as to provide richer and more accurate information for the early clinical diagnosis of breast cancer. Goal(s): To investigate the application value of semi-quantitative parameters of three enhanced sequences of MRI techniques in predicting Ki-67 expression in breast cancer. Approach: The semi-quantitative parameters of the enhanced images of the three groups were calculated respectively. The predictive parameters of Ki-67 expression in breast cancer were obtained by statistical analysis. Results: ROI1+54-per-brevity of enhancement was valuable for predicting the expression of Ki-67 in breast cancer(P=0.032). Impact: The DISCO-MRI with fast
scanning speed and high time resolution needs to be further studied whether it
can replace the traditional DCE-MRI scanning in the future, and it also needs
to find a suitable post-processing mode. |