EP Analytics (United States)

Abstract Dieser Beitrag stellt Handlungsfelder und Maßnahmen zur Verbesserung der Antwortqualität von Chatbots für die Produktion vor. Auf Basis einer explorativen Studie werden typische Fehlermuster von Chat-bots für die Produktion analysiert, drei Handlungsfelder identifiziert und Maßnahmen vorgestellt, mit denen die Antwortqualität gesteigert werden kann. Ein industrieller Anwendungsfall demonstriert die Wirksamkeit des Ansatzes.

Abstract This paper presents the dynamical core of the Climate Modeling Alliance (CliMA) atmosphere model, designed for efficient simulation of a wide range of atmospheric flows across scales. The core uses the nonhydrostatic equations of motion for a deep atmosphere, discretized with a hybrid approach that combines a spectral element method (SEM) in the horizontal and a staggered finite‐difference method in a height‐based, terrain‐following coordinate in the vertical. This approach leverages the high‐order accuracy and scalability of the SEM, while maintaining the computational efficiency and stability of finite differences on a staggered grid. The model's coordinate‐independent equation set allows for simulations in a variety of geometries and planetary configurations. The use of the specific total energy of moist air as a prognostic variable, along with a consistent thermodynamic formulation, ensures the conservation of energy, air mass, and water mass, even in moist atmospheres and in the presence of subgrid‐scale parameterizations, without ad hoc fixers. A horizontally explicit, vertically implicit (HEVI) timestepping strategy treats fast vertical processes implicitly and further enhances computational efficiency by allowing larger timesteps. The model demonstrates excellent strong and weak scaling on CPUs and GPUs, making it well‐suited for high‐resolution simulations on modern supercomputing architectures, including those on the cloud, which widens access to climate models.

The role of intragenic cytosine methylation in shaping phenotypes has been contentious. Recent studies show association between stress and alternative splicing of transcripts, but without functional genome-wide or single-position analysis. We utilized the msh1 experimental system in Arabidopsis as a model of reproducible epigenetic states with stress-responsive phenotypes, including commitment to heritable memory for at least seven generations. We mapped the methylome to single-cytosine resolution with signal-detection, verified by machine learning. Differentially methylated genes were overlapped with msh1-derived transcript isoforms to show that different patterns of exonic methylation led to different levels of isoform expression. Alternatively spliced and differentially methylated genes were enriched in key regulators of growth and development and spliceosome components. Genes targeted for differential methylation also contained a known CTT motif. These results demonstrate a direct relationship in plants between environmentally responsive differential methylation and alternative splicing behavior leading to phenotype changes.

Abstract Early Adoption of Molecular Residual Disease Testing in Breast Cancer Patients using Real World Data Authors: Ning Yan Gu1, Kohli M2, Stephen J Noga1, Quang A. Le1, Rodriguez-Silva J2, Barhoush Y2, Batchu L2, Gebra Cuyun Carter11Exact Sciences, Madison, WI, USA; 2STATinMED LLC., Dallas, TXWord count: 2765/3400 characters OBJECTIVE: Testing for molecular residual disease (MRD) via circulating tumor DNA (ctDNA) biomarker is an emerging practice in cancer treatment and disease recurrence detection. This study aims to assess the early adoption of MRD for patients diagnosed with breast cancer (BC) using real-world data. METHODS: A national multi-payer medical and pharmacy claims data (STATinMED RWD Insights) with nearly 80% of the US healthcare system was used. Data were extracted for female patients aged ≥18 years between January 1, 2015, and November 30, 2023. Index date was defined as the first date of BC diagnosis. Included patients were required to have continuous enrollment for a minimum of 12 months pre-index, 30 days post-index, and had ≥1 inpatient or ≥2 outpatient BC claims (ICD-9-CM: 174x, 175x; ICD-10-CM: C50x). Patients were excluded if they had metastasized disease, other cancers (not including non-melanoma skin cancers), any BC treatment, or had evidence of pregnancy during the 12-month baseline period. Proxies were used to determine BC subtype and cancer stage. MRD utilization was identified from the index date to study end time and/or patients’ death. Multivariate logistic regression was used to identify predictors of the MRD test adoption. RESULTS: Among 652,463 eligible patients (mean age: 65.4 years), 815 (0.12%) had evidence of MRD testing, the first one was in 2021, increased in 2022, and continued to the study end date of November 30th, 2023, with most of the MRD testing occurred in 2023 (n=748, 91.8%). The mean follow-up time per patient was 38.76 months (SD: 26.57) with a median follow-up per patient of 29.70 months. The mean age of MRD tested patients was 62.7 years; 67.7% were White; 60.1% were non-Hispanic; 36.0% had a college education; 51.4% had Medicare; 27.3% had an income below $25,000 and 55.5% resided in the Southern United States. Categorizing the subtype, 65.2% had HR+/HER2- BC, 23.3% had triple-negative BC (TNBC), and 9.5% had HR+/HER2+ BC. During the study time frame, 60.9% of patients received only one MRD test, 23.1% received two, 8.3% received three and the remaining 7.7% of patients received four or more MRD tests, for up to 10 MRD tests. The mean number of MRD tests per patient was 1.72 (SD: 1.27). Among the 815 MRD-tested patients, 392 (48%) had evidence of breast-conserving surgery (BCS). Of those 392 patients, 5 (1.3%) received MRD testing before the BCS and 387 (98.7%) received MRD test after the BCS. For every BC subtype, the average time between tests was > 80 days between the 1st and 2nd tests, > 70 days between the 2nd and 3rd tests, with shorter intervening times for additional MRD tests conducted during study period. Based on the multivariate logistic regression, key factors associated with the receipt or adoption of MRD test included younger age, advanced BC stage, BC subtype, and patients’ residence region. CONCLUSION: This descriptive analysis using real-world data offers insights into the early adoption of the MRD testing among BC patients in the United States. The real-world evidence showcased an upward adoption trajectory during the study period and there were variations in early utilization patterns. Findings herein provide hypothesis generating opportunities for future investigations. Key Words: 1. Molecular residual disease 2. MRD testing 3. ctDNA 4. Breast cancer 5. Real world data Citation Format: N. Gu, M. Kohli, S. J. Noga, Q. A. Le, J. Rodriguez-Silva, Y. Barhoush, L. Batchu, G. C. Carter. Early Adoption of Molecular Residual Disease Testing in Breast Cancer Patients using Real World Data [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 PS2-04-01.

This study evaluates pool boiling on additively manufactured copper surfaces with various microstructures, using distilled water under saturated atmospheric conditions. Initially, heat-treated and untreated samples were compared to assess thermal conductivity effects. Heat-treated samples, despite higher thermal conductivity, generally showed lower heat transfer coefficients (HTC) due to smoother surfaces and fewer active nucleation sites. Further testing involved heat-treated surfaces with channels, tunnels, chimneys, and pillars of varying heights, benchmarked against a flat surface. Chimney structures achieved the highest enhancements, surpassing 3000 kW m −2 in maximum heat flux and an HTC of 260 kW m −2 K −1 , which is a 400 % improvement compared to the reference. Their superior performance resulted from efficient liquid-vapor separation, capillary wicking, and favorable bubble dynamics facilitated by their geometry. Pillar structures significantly enhanced critical heat flux but had limited HTC due to vapor entrapment and bubble coalescence. In contrast, chimney features provided balanced boiling performance across diverse heat fluxes. Overall, this study demonstrates the promise of laser powder bed fusion to create advanced copper surfaces for effective thermal management applications, particularly in systems demanding high heat dissipation, minimal surface superheat, and complex geometries. • Additive manufacturing enabled efficient vapor-liquid separation via custom geometries. • Chimney microfeatures achieved a 400 % heat transfer coefficient enhancement. • Heat-treated copper surfaces enhanced CHF up to 3065 kW·m-2 in boiling experiments. • Pillar geometries improved CHF but limited HTC due to vapor entrapment. • Study reveals structure-specific design guidelines for boiling heat transfer optimization.

The growing number of global shocks, which include geopolitical instabilities, pandemics, and climate shock, among others, has revealed major weaknesses in the U.S. manufacturing supply chains. The need to increase supply chain resiliency and, at the same time, improve sustainability and resource efficiency has become a strategic necessity. The proposed study is an AI- and machine learning-controlled predictive analytics-based framework created to enhance the reliability and resilience of the U.S. manufacturing supply chain systems. The framework incorporates heterogeneous data sources such as operational, environmental, market, and logistics data to facilitate the proactive identification of risks, prediction of demand, and optimization of resources. Developed machine learning models are used to embrace non-linear relationships with complexities across the supply chain nodes to assist in real-time decision-making and adaptive response strategies. The proposed framework enables the identification of disruption early in the supply chain planning and execution, better management of inventory and capacity, reduced waste, and increased energy and material efficiency by integrating predictive analytics into the supply chain planning and execution. In theory, this study will be a step forward in the interplay of artificial intelligence, sustainable supply chain management, and resilience engineering. In practice, it provides decision-makers with a scalable and evidence-based instrument in order to enhance the continuity of operations, environmental performance, and competitiveness in the long run of the U.S. manufacturing supply chains.

A 48-year-old man presented to a rural Honduran hospital with progressive dyspnea for three days, worsening at rest. He was hemodynamically stable and without tachypnea or pulmonary crackles. The electrocardiogram showed sinus tachycardia, suggestive of pulmonary thromboembolism (PTE). Chest CT confirmed the diagnosis of bilateral PTE with total occlusion of the right pulmonary artery and significant stenosis of the left pulmonary artery. Transthoracic echocardiogram revealed a right atrial thrombus. Medical treatment with therapeutic anticoagulation was provided. Echocardiographic follow-up at 10 days showed that the original thrombus had been significantly reduced to two smaller thrombi. The patient was discharged from the hospital with more anticoagulants. Thrombus in transit alongside pulmonary embolism is rare and is an emergent condition; it can quickly progress to cardiogenic shock or cardiac arrest. This patient’s effective diagnosis and treatment highlight the importance of bedside echocardiography in resource-limited settings.

This study re-examined the assumption that promotion systems and advanced HR technologies operate under neutral meritocratic principles. It explored how formal promotion criteria, algorithmic scoring tools, and generative-AI screening shaped long-term organizational productivity, with particular attention to perceptions of fairness, transparency, and legitimacy. A mixed-methods design was employed, combining survey data from employees, managers, and HR professionals with qualitative interviews from AI-enabled organizations. Quantitative analysis revealed that although algorithmic and AI-driven HR processes increased administrative efficiency and consistency, these technologies did not independently predict higher productivity. Instead, perceived fairness and transparency emerged as significantly stronger predictors of employee engagement and productivity outcomes. Regression analysis further indicated that fairness perceptions exerted the largest influence on productivity compared with algorithmic scoring and AI screening mechanisms. Qualitative findings supported these results, showing that employees were more likely to accept AI-mediated decisions when clear explanations, human oversight, and accessible appeals processes were present. Conversely, opaque or fully automated systems were associated with distrust, perceived bias, and weaker motivational outcomes. The study concluded that AI should complement, rather than replace, human decision-makers and that responsible governance frameworks were necessary to ensure equity and sustained performance benefits. Recommendations focused on transparency design, hybrid decision structures, bias audits, and digital literacy development. Future research was suggested in cross-cultural contexts, longitudinal adoption effects, and explainable-AI applications in HR decision-making.

A key factor in increasing the efficiency of perovskite solar cell (PCSs) is optimizing the thickness of the absorber and the operating temperature. In order to examine the impact of these factors, a planar n-i-p PSCs with the configuration FTO/α-Fe 2 O 3 /MAPbI 3 /Spiro-OMeTAD/Au was simulated using SCAPS-1D. The results indicate that the charge carrier generation and transportation are strongly governed by the thickness of MAPbI3 absorber layer. When tune to 600 nm thickness, the device has a Jsc of 22.1 mA/cm 2 , an Voc of 1.06 V, a FF of 77.4 %, and an overall PCE of 18.3 %. When layers are too thin (<300 nm), they limit the photon absorption, which lower the Jsc. When layers are too thick (>800nm), they cause recombination losses that lower the FF. Variation in the temperature also affect how well the device works in different temperature conditions. For example, raising form 300 K to 350 K lowers the Voc and PCE because of non-radiative recombination happens more often. These finding highlight that careful control of absorber thickness coupled with thermal stability of the α-Fe2O3 ETL is crucial for achieving stable and high-efficiency PSCs.