Search

PURPOSE: To develop an efficient deep neural network model that incorporates context from neighboring image sections to detect breast cancer on digital breast tomosynthesis (DBT) images. MATERIALS AND METHODS: The authors adopted a transformer architecture that analyzes neighboring sections of the DBT stack. The proposed method was compared with two baselines: an architecture based on three-dimensional (3D) convolutions and a two-dimensional model that analyzes each section individually. The models were trained with 5174 four-view DBT studies, validated with 1000 four-view DBT studies, and tested on 655 four-view DBT studies, which were retrospectively collected from nine institutions in the United States through an external entity. Methods were compared using area under the receiver operating characteristic curve (AUC), sensitivity at a fixed specificity, and specificity at a fixed sensitivity. RESULTS: On the test set of 655 DBT studies, both 3D models showed higher classification performance than did the per-section baseline model. The proposed transformer-based model showed a significant increase in AUC (0.88 vs 0.91, P = .002), sensitivity (81.0% vs 87.7%, P = .006), and specificity (80.5% vs 86.4%, P < .001) at clinically relevant operating points when compared with the single-DBT-section baseline. The transformer-based model used only 25% of the number of floating-point operations per second used by the 3D convolution model while demonstrating similar classification performance. CONCLUSION: A transformer-based deep neural network using data from neighboring sections improved breast cancer classification performance compared with a per-section baseline model and was more efficient than a model using 3D convolutions. Keywords: Breast, Tomosynthesis, Diagnosis, Supervised Learning, Convolutional Neural Network (CNN), Digital Breast Tomosynthesis, Breast Cancer, Deep Neural Networks, Transformers Supplemental material is available for this article. © RSNA, 2023

Moyamoya disease, a rare chronic cerebrovascular disease with a fragile vascular network at the base of the brain, can cause ischemic or hemorrhagic strokes or seizures. Precise blood pressure control and adequate analgesia are important for patients with moyamoya disease to prevent neurological events such as ischemia and hemorrhage. This study aimed to compare the intraoperative mean arterial pressure of pregnant women with moyamoya disease according to the mode of anesthesia (general anesthesia versus spinal anesthesia) used during cesarean delivery.We retrospectively reviewed the medical records of 87 cesarean deliveries in 74 patients who had been diagnosed with moyamoya disease before cesarean delivery. The primary outcome, intraoperative maximum mean arterial pressure during anesthesia, was compared according to the type of anesthesia administered (general versus spinal anesthesia). Other perioperative hemodynamic data (lowest mean arterial pressure, incidence of hypotension, vasopressor use, and antihypertensive agent use), maternal neurologic symptoms, neonatal outcomes (Apgar scores7, ventilatory support, and intensive care unit admission), maternal and neonatal length of stay, postoperative pain scores, and rescue analgesic use were assessed as secondary outcomes.While the lowest blood pressure during anesthesia and incidence of hypotension did not differ between the 2 groups, the maximum mean arterial pressure during anesthesia was lower in the spinal anesthesia group than that in the general anesthesia group (104.8 ± 2.5 vs 122.0 ± 4.6; P = .002). Study data did not support the claim that maternal neurologic symptoms differ according to the type of anesthesia used (5.6% vs 9.3%; P = .628); all patients recovered without any sequelae. The postoperative pain scores were lower, and fewer rescue analgesics were used in the spinal anesthesia group than in the general anesthesia group. Other maternal and neonatal outcomes were not different between the 2 groups.Compared with general anesthesia, spinal anesthesia mitigated the maximum arterial blood pressure during cesarean delivery and improved postoperative pain in patients with moyamoya disease.

Background Mammographic Artificial intelligence (AI) algorithms (Lunit insight MMG) draws attention as a diagnostic support tool for breast cancer detection. Lunit insight MMG provides a location suspected of breast cancer with a heatmap and a score reflecting the probability of the presence of suspicious areas. We investigated whether the Lunit insight MMG score is relevant for predicting the response to adjuvant tamoxifen. Methods Patients diagnosed with DCIS and underwent treatment at Seoul National University Hospital in 2010 were retrospectively enrolled. Clinical characteristics, tamoxifen use, survival data, and mammography images were extracted from the electronic medical records, and Lunit insight MMG scores were calculated retrospectively. We classified two groups according to tamoxifen treatment and compared the score change of contralateral breast from baseline to 5 years after surgery for DCIS. Change categories of Lunit insight MMG score included maintaining high risk, maintaining low risk, increasing from low risk to high risk, and decreasing from high risk to low risk. Results Of 100 patients, 50 (50%) had undergone tamoxifen treatment (group 1) and 50 (50%) had not (group 2). The median age of the patients was 48.4 years for group 1 and 51.5 years for group 2 (p=0.172). The median follow-up duration was 8.7 years for the whole cohort. Using Lunit insight MMG score, more patients in group 1 decreased in contralateral breast cancer (CBC) risk compared with group 2 (6.0% vs. 2.0%; P=.008). No patients in group 1 had an increase in CBC risk while 9 patients increased in group 2 (0% vs. 18.0%; P =.008). There was no ipsilateral breast cancer recurrence for the whole cohort, and two patients experienced contralateral invasive breast cancer in group 2. In two patients with CBC, the Lunit insight MMG score increased five years after surgery, one year and three years before the CBC diagnosis. Conclusions Longitudinal Changes of Mammographic AI algorithms Score may be a predictive surrogate marker for response to tamoxifen therapy in hormone receptor-positive DCIS. Citation Format: Changjin Lim, Eun Kyung Park, Hong-Kyu Kim, Eunhye Kang, Ji-Jung Jung, Hyunsu Yeoh, Jang-il Kim, Jung Whan Chun, Han-Byoel Lee, Hyeong-Gon Moon, Wonshik Han. Longitudinal Changes of Contralateral Breast Mammographic Artificial Intelligence Algorithms Score in Ductal Carcinoma In Situ patient with Tamoxifen [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P3-04-11.

PURPOSE The purpose of this study was to identify the optimal timing for screening spinal cord ultrasonography (US) to detect filum terminale lipoma in infants. METHODS We retrospectively reviewed infants (

This paper highlights the role of local institutions and innovation systems in supporting the participation and upgrading of firms in the global value chain through a case study of the export-oriented fresh fruit and vegetable sector in Kenya. The case illustrates how sectoral institutions at the local and global levels create a connection between the global value chain and local innovation systems and facilitate local learning that is necessary for supplying global buyers in advanced markets. KenyaGAP is a formal national institution translating global certification procedures into the local setting, thereby providing learning opportunities for small Kenyan exporters to become acquainted with the quality standards required by foreign buyers. This case demonstrates that the timely establishment of local institutions and public–private interactions in the process of developing local institutions has been critical in supporting interactive learning within local innovation systems.

This article shows evidence that FDI policies during the catch-up process may leave a trace in the openness of innovation activities in latecomer economies, based on a comparative analysis between the Republic of Korea and China. The past industrial policies of the Republic of Korea favoured creating local technological competence based on the transfer of foreign knowledge in codified form, leading to a low level of global connection in local knowledge creation. By contrast, Chinese policies encouraged the entrance of foreign firms in the Chinese market, leading to a higher level of global interaction in innovation activities. Based on the findings, the article presents policy recommendations and suggests avenues for future research.

Lithium (Li) metal is considered an ideal anode for next-generation Li metal batteries (LMBs). However, Li dendrite formation and low coulombic efficiency (CE) are major obstacles for its practical use in LMBs. The engineering of the interface between Li metal and electrolytes is crucial for better Li ion kinetics and interface stability to overcome the above problems. In this study, we report a thin LiF/defluorinated polymer (LiF@Po) composite as an artificial SEI layer. The LiF@Po composite layer, formed by a facile and scalable roll-pressing process using polytetrafluoroethylene (PTFE) and Li metal, is composed of an inner (lithium side) LiF-rich layer and an outer (electrolyte side) polymer rich (polyene/fluoropolymer) layer. The free space between LiF particles could be conformally filled with a polymeric layer, improving the Li surface coverage properties of the protection layer. This novel LiF@Po protection layer enables highly stable cyclability even in a carbonate-based electrolyte for more than 1000 h at 1 mA cm−2 in a Li‖Li symmetric cell and a cycle retention of 80% capacity for >200 cycles with an extremely stable average CE of 99.7% in a Li‖LiCoO2 full cell. Furthermore, this LiF@Po protection layer could significantly enhance the cyclability and reversible capacity in a Li‖LiNi0.8Co0.1Mn0.1O2 (NCM-811) full cell.

In this chapter we analyse the green transition of the Danish national system of innovation with a focus on green patents, green skills and green innovations. We find that green patenting, green skills and green innovations are inter-connected in the green transition. The Danish technological profile has shown an increase in green patents, and green innovation now happens in all parts of Denmark. We find that the education and training system must be adapted if Danish firms are to lead in the green transition. The findings point to challenges, requirements and opportunities created by the transition to a greener economy.

Objective Low-grade epilepsy-associated tumors (LEATs) are associated with childhood seizures that are typically drug-resistant, necessitating surgical interventions. In this study, we aimed to investigate the efficacy of surgical intervention in children with LEATs and to identify factors associated with seizure and cognitive outcomes. Methods We reviewed 58 children less than 18 years of age who underwent epilepsy surgery due to histopathologically confirmed LEATs and had a minimum postoperative follow-up duration of 24 months. Results Of the 58 patients who were followed for a median duration of 5.6 (IQR 3.2 to 10.0) years, 51 (87.9%) were seizure-free after surgery. In univariate analysis, shorter epilepsy duration, fewer antiepileptic drugs at time of surgery, gross total resection, and unilobar tumor involvement were associated with seizure freedom. In multivariate analysis, gross total resection was independently associated with seizure freedom. The preoperative and postoperative full-scale intelligence quotient (FSIQ) scores were 78.9 ± 27.1 and 80.9 ± 28.7, respectively. In univariate analysis, younger age at seizure onset, longer epilepsy duration, more antiepileptic drugs at time of surgery, multilobar tumor involvement, and presence of generalized epileptic discharges were associated with lower preoperative FSIQ. In multivariate analysis, longer epilepsy duration was independently associated with lower preoperative FSIQ scores. Postoperative FSIQ scores were significantly influenced by preoperative FSIQ scores. Conclusions Epilepsy surgery for LEATs in children resulted in excellent seizure outcome. Gross total resection was the only independent factor associated with favorable seizure outcome. Preoperative and postoperative cognitive abilities were significantly influenced by epilepsy duration, so early surgical intervention should be considered.

10533 Background: Accurate risk assessment allows the precise personalized screening of breast cancer. Conventional statistical risk models have been used that estimate the probability of breast cancer incidence with patient demographics, detailed personal and family history. The purpose of this study was to investigate the feasibility of the AI-powered Imaging Biomarker in mammography (IBM) which was developed to predict the risk of future breast cancer beyond the mammographic breast density. Methods: We trained and developed the deep learning risk model, the AI-powered IBM, using a total of 36,955 examinations from 21,438 patients, who underwent at least one mammogram using Hologic or Siemens machines and pathology-confirmed breast cancer outcomes. To discover the feasibility of the AI-powered IBM, mammograms and various clinical information including pathology-confirmed breast cancer outcomes were collected, which were only used for external validation. C-indices and receiver operating characteristic (ROC) curves for 1- to 5-year outcomes were obtained. We compared 5-year ROC area under the curves (AUCs) of our AI-powered IBM and statistical risk models including the Tyrer-Cuzick model and the Gail model, which were most commonly used and widely known, using DeLong’s test. Results: A total of 16,894 mammograms were collected for external validation, of which 4,002 were followed by a cancer diagnosis within 5 years. Our AI-powered IBM obtained C-index of 0.758, and the model demonstrated the risk of breast cancer with AUC of 0.895 (95% CI: 0.880, 0.909) at 1-year, 0.839 (0.824, 0.852) at 2-year, 0.807 (0.794, 0.819) at 3-year, 0.783 (0.7947, 0.819) at 4-year. The 5-year AUC of our AI-powered IBM was 0.808 (0.792, 0.822). Our AI-powered IBM showed significantly higher 5-year AUC than the Gail model (AUC: 0.572, P< 0.001) and the Tyrer-Cuzick model (0.569, P< 0.001). Conclusions: A deep learning AI-powered IBM using mammograms has a substantial potential to advance toward the robust risk prediction of breast cancer over conventional risk models. This approach for risk stratification of breast cancer might be feasible to improve personalized screening programs.

Objective With the recognition of epilepsy as a network disease that disrupts the organizing ability of resting-state brain networks, vagus nerve stimulation (VNS) may control epileptic seizures through modulation of functional connectivity. We evaluated preoperative 2-deoxy-2[18 F]fluoro-D-glucose (FDG) positron emission tomography (PET) in VNS-implanted pediatric patients with refractory epilepsy to analyze the metabolic connectivity of patients and its prognostic role in seizure control. Methods Preoperative PET data of 66 VNS pediatric patients who were followed up for a minimum of 1 year after the procedure were collected for the study. Retrospective review of the patients' charts was performed, and five patients with inappropriate PET data or major health issues were excluded. We conducted an independent component analysis of FDG-PET to extract spatial metabolic components and their activities, which were used to perform cross-sectional metabolic network analysis. We divided the patients into VNS-effective and VNS-ineffective groups (VNS-effective group, ≥50% seizure reduction; VNS-ineffective group, Results Thirty-four (55.7%) patients showed >50% seizure reduction from baseline frequency 1 year after VNS. A significant difference in metabolic connectivity evaluated by preoperative FDG-PET was noted between groups. Relative changes in glucose metabolism were strongly connected among the areas of brainstem, cingulate gyrus, cerebellum, bilateral insula, and putamen in patients with Significance This study shows that seizure outcome of VNS may be influenced by metabolic connectivity, which can be obtained from preoperative PET imaging. This study of metabolic connectivity analysis may contribute in further understanding of the mechanism of VNS in intractable seizures.

To investigate the feasibility and clinical effectiveness of performing multiple burr hole surgery in pediatric moyamoya patients as a response to failed modified encephaloduroarteriosynangiosis (mEDAS). From January 2014 to May 2018, multiple burr hole surgery (MBS) was conducted on 16 hemispheres in 12 patients as a secondary treatment following mEDAS. The male-to-female ratio was 1:2 and the average age at the time of mEDAS was 6 years old. The average patient age was 9 ± 3 years olds (range 7–17) at the time of MBS which occurred an average of 46 months after mEDAS. An average of 10 ± 1 holes (range 8–13) were made. Time-to-peak (TTP) magnetic resonance images (MRI) were taken along 20 axial cuts. Of these cuts, two consecutive cuts on the lateral ventricle were selected to calculate the average value of the region of interest (ROI). The value of the cerebellum was subtracted from the average value of two consecutive cuts. The ROI value was analyzed using a paired t test by SPSS 20 (SPSS Inc., Chicago, IL, USA). All 16 cases presented improvement of clinical symptoms as determined by ROI analysis of the TTP MRI images. The average ROI value was 5.03 ± 6.36 before MBS and − 15.54 ± 9.42 after MBS. The average change in the ROI value was − 20.58 ± 12.59. The ROI value decreased in all cases after MBS. Magnetic resonance angiography (MRA) also showed a positive effect on vascularization. In pediatric moyamoya patients, MBS is recommended as secondary option as a response to failed mEDAS. Its clinical effectiveness was shown by analyzing TTP images and assisted by MRA and digital subtraction angiography.

Objectives: We evaluated the effects of a comprehensive antimicrobial stewardship program (ASP) in a surgical intensive care unit (SICU). Methods: The ASP was implemented from March 2018 to February 2019 at an SICU in a teaching hospital. An infectious disease physician and a pharmacist visited the SICU 3 times per week for prospective audit and feedback. Outcomes were compared between the ASP period and the same months in the preceding year (pre-ASP period). The primary outcome measure was the use of anti-pseudomonal beta-lactams (APBL). Appropriate antimicrobial de-escalation and ICU mortality rates were also compared. Results: A total of 182 and 149 patients were included in the study for the pre-ASP and ASP periods, respectively. Although disease severity was higher in the ASP group (septic shock 39.0% in pre-ASP vs 65.1% in ASP group, P

The literature on entrepreneurial ecosystem places emphasis on understanding the surrounding business environment for entrepreneurial activities of individual actors in a system, typically at the city or regional level. While gaining prominence as a useful concept to provide guidance for innovation and entrepreneurship policy, the framework at its current stage faces conceptual as well as empirical challenges with regards to the structural boundary of the ecosystem and measurement. This chapter aims to jointly address these issues by deploying a novel combination of methods and data sources to map the interaction within a dynamic emerging ecosystem in Nairobi, Kenya. We do so by identifying start-ups and investors utilising data from CrunchBase and exploring the dynamics of the relevant actors in the Twitter network, where we use manual categorisation as well as data-driven community detection algorithms to identify social networks in the ecosystem. We identify distinct communities in the ecosystem based on the technological focus, the types of support organisations, and interaction patterns in the network. We show how technology start-ups in developing economies are connected to support organisations of various geographical origins, which implies that ecosystems for supporting entrepreneurship can emerge across borders.

Background Shear wave elastography (SWE) is an ultrasound technique for the noninvasive quantification of tissue stiffness. The hypoxic tumor microenvironment promotes tumor stiffness and is associated with poor prognosis in cancer. We aimed to investigate the correlation between tumor hypoxia and histologic biomarkers and tumor stiffness measured by SWE in breast cancer. Methods From June 2016 to January 2018, 82 women with invasive breast cancer who underwent SWE before treatment were enrolled. Average tumor elasticity (Eaverage) and tumor-to-fat elasticity ratio (Eratio) were extracted from SWE. Immunohistochemical staining of glucose transporter 1 (GLUT1) was used to assess tumor hypoxia in breast cancer tissues and automated digital image analysis was performed to assess GLUT1 activities. Spearman correlation and logistic regression analyses were performed to identify associations between GLUT1 expression and SWE values, histologic biomarkers, and molecular subtypes. The Mann–Whitney U test, t test, or Kruskal–Wallis test was used to compare SWE values and histologic features according to the GLUT1 expression (≤the median vs > median). Results Eaverage (r = 0.676) and Eratio (r = 0.411) correlated significantly with GLUT1 expression (both p average was significantly higher in cancers with estrogen receptor (ER)–, progesterone receptor (PR)–, Ki67+, and high-grade (p ratio was higher in cancers with Ki67+, lymph node metastasis, and high-grade (p median) had higher Eaverage (mean, 85.4 kPa vs 125.5 kPa) and Eratio (mean, 11.7 vs 17.9), and more frequent ER– (21.7% vs 78.3%), PR– (26.4% vs 73.1%), Ki67+ (31.7%% vs 68.3%), human epidermal growth factor receptor 2 (HER2) + (25.0% vs 75.0%), high-grade (28.6% vs 71.4%), and HER2-overexpressing (25.0% vs 75.0%) and triple-negative (23.1% vs 76.9%) subtypes (p average was independently associated with GLUT1 expression (p Conclusions Tumor stiffness on SWE is significantly correlated with tumor hypoxia as well as histologic biomarkers. In particular, Eaverage on SWE has independent prognostic significance for tumor hypoxia in the multivariable analysis and can potentially be used as a noninvasive imaging biomarker to predict prognosis and pretreatment risk stratification in breast cancer patients.

To provide the insight for postoperative hypotonia. Selective posterior rhizotomy (SPR) has been proved as a powerful tool for reducing spasticity. And also, its functional benefit and long-term effect are also well-known.The most considered side effect of this procedure is postoperative hypotonia. However, some extent of temporary postoperative hypotonia can be the marker of the long-term success of this procedure. While the return of spasticity is the most unwanted side effect, some kind of overfitting, temporary postoperative hypotonia, can be the solution for that.For severely deformed patients, postoperative hypotonia may not be problematic, because severe spasticity makes them deformed and disabled. Deformed body will not show a definite disability from postoperative hypotonia.

Background and purpose Moyamoya disease (MMD) is a rare form of intracranial stenoocclusive disease that can be associated with intracranial aneurysms. We evaluated the clinical features and outcomes of MMD-associated aneurysms while focusing on their locations. Methods Between January 1998 and December 2018 there were 1,302 adult and pediatric patients diagnosed as MMD at a single institution. These patients included 38 with 44 MMD-associated aneurysms. The MMD-associated aneurysms were classified into two groups based on their locations: major-artery aneurysms and non-major-artery aneurysms. The clinical and radiological data for patients with MMD-associated aneurysms were reviewed retrospectively. Results The 44 MMD-associated aneurysms comprised 28 in major arteries and 16 in nonmajor arteries. All of the major-artery aneurysms were initially unruptured lesions, and follow-up angiography showed that 23 (82.1%) had an improved or stable status and 5 (17.9%) had a worse status. The non-major-artery aneurysms comprised 10 ruptured and 6 unruptured lesions, and follow-up angiography showed that 11 (68.8%) had improved or were stable and 5 (31.2%) had worsened. At the latest follow-up, there were four cases of unfavorable outcome: two initial hemorrhagic insults, one treatment-related morbidity, and one repeated-hemorrhage case. Conclusions MMD-associated aneurysms occurred in 3.3% of the MMD cohort in this study, of which 63.6% were major-artery aneurysms and 36.4% were non-major-artery aneurysms. The major-artery group included 17.9% that became angiographically worse, while 31.2% were growing or hemorrhaging in the non-major-artery group.

Background Computer-aided detection (CAD) can detect breast lesions by using an enhancement threshold. Threshold means the percentage of increased signal intensity in post-contrast imaging compared to precontrast imaging. If the pixel value of the enhanced tumor increases above the set threshold, CAD provides the size of the tumor, which is calculated differently depending on the set threshold. Therefore, CAD requires the accurate setting of thresholds. We aimed to compare the diagnostic accuracy of tumor size measurement using MRI and CAD with 3 most commonly used thresholds and to identify which threshold is appropriate on CAD in breast cancer patients. Methods A total of 130 patients with breast cancers (80 invasive cancers and 50 ductal carcinoma in situ [DCIS]) who underwent preoperative MRI with CAD and surgical treatment were included. Tumor size was manually measured on first contrast-enhanced MRI and acquired by CAD using 3 different thresholds (30, 50, and 100%) for each tumor. Tumor size measurements using MRI and CAD were compared with pathological sizes using Spearman correlation analysis. For comparison of size discrepancy between imaging and pathology, concordance was defined as estimation of size by imaging within 5 mm of the pathological size. Concordance rates were compared using Chi-square test. Results For both invasive cancers and DCIS, correlation coefficient rho (r) between tumor size on imaging and pathology was highest at CAD with 30% threshold, followed by MRI, CAD with 50% threshold, and CAD with 100% threshold (all p p = 0.213). For DCIS, the concordance rate of 30.0% at CAD with 30% threshold showed no difference with that of 36.0% at MRI (p = 0.699). Compared to MRI, higher risk of underestimation was noted when using CAD with 50% or 100% threshold for invasive cancers and when using CAD with 100% threshold for DCIS. Conclusion For CAD analysis, 30% threshold is the most appropriate threshold whose accuracy is comparable to manual measurement on MRI for tumor size measurement. However, clinicians should be aware of the higher risk of underestimation when using CAD with 50% threshold for tumor staging in invasive cancers.

Background Radiogenomic investigations for breast cancer provide an understanding of tumor heterogeneity and discover image phenotypes of genetic variation. However, there is little research on the correlations between US features of breast cancer and whole-transcriptome profiling. Purpose To explore US phenotypes reflecting genetic alteration relevant to breast cancer treatment and prognosis by comparing US images of tumor with their RNA sequencing results. Materials and Methods From January to October 2016, B-mode and vascular US images in 31 women (mean age, 49 years ± 9 [standard deviation]) with breast cancer were prospectively analyzed. B-mode features included size, shape, echo pattern, orientation, margin, and calcifications. Vascular features were evaluated by using microvascular US and contrast agent-enhanced US: vascular index, vessel morphologic features, distribution, penetrating vessels, enhancement degree, order, margin, internal homogeneity, and perfusion defect. RNA sequencing was conducted with total RNA obtained from a surgical specimen by using next-generation sequencing. US features were compared with gene expression profiles, and ingenuity pathway analysis was used to analyze gene networks, enriched functions, and canonical pathways associated with breast cancer. The P value for differential expression was extracted by using a parametric F test comparing nested linear models. Results Thirteen US features were associated with various patterns of 340 genes (P < .05). Nonparallel orientation at B-mode US was associated with upregulation of TFF1 (log twofold change [log2FC] = 4.0; P < .001), TFF3 (log2FC = 2.5; P < .001), AREG (log2FC = 2.6; P = .005), and AGR3 (log2FC = 2.6; P = .003). Complex vessel morphologic structure was associated with upregulation of FZD8 (log2FC = 2.0; P = .01) and downregulation of IGF1R (log2FC = -2.0; P = .006) and CRIPAK (log2FC = -2.4; P = .01). The top networks with regard to orientation or vessel morphologic structure were associated with cell cycle, death, and proliferation. Conclusion Compared with RNA sequencing, B-mode and vascular US features reflected genomic alterations associated with hormone receptor status, angiogenesis, or prognosis in breast cancer. © RSNA, 2020 Online supplemental material is available for this article.