154092 research outputs found
Sort by
The late Middle Pleistocene Homo erectus of the Madura Strait, first hominin fossils from submerged Sundaland
Get PDFEastern Asia yielded a rich fossil record of Pleistocene hominins, ranging from Homo erectus and the diminutive island species Homo floresiensis and Homo luzonensis, to post-erectus grade late archaic Homo (including Denisovans), and finally to anatomically modern humans. The Sunda Shelf played an important role in the dispersal and evolution of hominin populations. The shelf has been widely exposed during most of the Pleistocene, forming a landmass known as Sundaland. Today, the area holds the world’s largest shelf sea. Thus far, hominin fossils from submerged Sundaland were not available. Here we report on the finding of two hominin cranial fragments from the submerged Sunda Shelf, retrieved during a dredging work in the Madura Strait, off the Java coast. The specimens derive from the sandy fill of a late Middle Pleistocene submerged valley of the Solo River and consist of a frontal fragment and a parietal fragment. Metric and morphological comparisons with Pleistocene skulls from the Asian mainland, Java and Flores point to a relation with the late Homo erectus of Java, in particular with the crania from Sambungmacan. The Madura Strait hominins were probably part of an MIS6 population that lived along the Solo, which in this period continued eastward over the exposed shelf area of the Madura Strait. Probably, the large perennial rivers of Sundaland offered good living conditions for Homo erectus, in a late Middle Pleistocene climate setting that was relatively dry
Conceptual modeling:Foundations, a historical perspective, and a vision for the future
Get PDFWe recount the foundations of Conceptual Modeling in Computer Science, Philosophy and Cognitive Science and their implications on what are concepts, conceptualizations, and conceptual models. We then review the history of the field, considering earlier work by the three co-authors, and highlight some of the contributions that made it what it is. Finally, we propose three research directions whose solutions could advance the field and will hopefully be addressed in the future. Our study is intended to help to circumscribe and characterize the field. It draws ideas from Philosophy, Cognitive Science, Engineering and the Social Sciences, as well as several areas within Computer Science, including Programming languages, Artificial Intelligence, Databases, Software Engineering, and Information Systems Engineering
Towards an open soil-plant digital twin based on STEMMUS-SCOPE model following open science
Get PDFDroughts and heatwaves jeopardize terrestrial ecosystem services. The development of an open digital twin of the soil-plant system can help monitor and predict the impact of these extreme events on ecosystem functioning. We illustrate how our recently developed STEMMUS-SCOPE model—STEMMUS, Simultaneous Transfer of Energy, Mass and Momentum in Unsaturated Soil; SCOPE, Soil Canopy Observation of Photosynthesis and Energy fluxes—links soil-plant processes to novel satellite observables (e.g. solar-induced chlorophyll fluorescence), contributing to such a digital twin. This soil-plant digital twin allows a mechanistic window for tracking above- and below-ground ecophysiological processes with remote sensing observations. Following Open Science and FAIR (Findable, Accessible, Interoperable, Reusable) principles, both for data and research software, we present the building blocks of the soil-plant digital twin. It emphasizes the importance of FAIR-enabling digital technologies to translate research needs and developments into reproducible and reusable data, software and knowledge
Enriching urban digital twins with energy-related Information from aerial and street view imagery for precise urban climate modeling
Get PDFThe country-scale 3D city model of the Netherlands, 3DBAG, has been extensively utilized across multiple disciplines. Buildings in 3DBAG include many attributes, but they lack the necessary ones required for precise urban climate modeling. This study aims to enhance the 3DBAG 3D city model by incorporating energy-related attributes and vegetation data, thereby improving visualization accuracy and realism, as well as enabling more precise urban climate simulations. Automated methods were developed to extract and integrate key features such as roof textures, façade materials, albedo values, roof slopes, and urban vegetation. Roof textures extracted from orthophotos enhance the analysis of roof geometry and facilitate material identification and solar panel detection. Meanwhile, façade materials and albedo values contribute to more accurate simulations of heat absorption, energy balance, and urban heat island effects. Furthermore, the integration of detailed urban vegetation data, including tree height and crown diameter, allows for accurate modeling of vegetation's influence on urban microclimates. Developed automated workflows significantly reduce the time and effort required for manual data preparation and integration. By enriching the urban digital twins with energy-related information, this study provides researchers with a more comprehensive and precise dataset, enabling more accurate analyses in urban climate modeling. The developed automated workflows can be applied in future releases of the 3DBAG, allowing other users in utilizing the energy-enriched 3D city model effectively. Web-based visualization of the energy-enriched 3D city model can be explored at https://bit.ly/4tuheritage.</p
Senmap:Multi-objective data-flow mapping and synthesis for hybrid scalable neuromorphic systems
Get PDFThis paper introduces SENMap, a mapping and synthesis tool for scalable, energy-efficient neuromorphic computing architecture frameworks. SENECA is a flexible architectural design optimized for executing edge AI SNN/ANN inference applications efficiently. To speed up the silicon tape-out and chip design for SENECA, an accurate emulator, SENSIM, was designed. While SENSIM supports direct mapping of SNNs on neuromorphic architectures, as the SNN and ANNs grow in size, achieving optimal mapping for objectives like energy, throughput, area, and accuracy becomes challenging. This paper introduces SENMap, flexible mapping software for efficiently mapping large SNN and ANN applications onto adaptable architectures. SENMap considers architectural, pretrained SNN and ANN realistic examples, and event rate-based parameters and is open-sourced along with SENSIM to aid flexible neuromorphic chip design before fabrication. Experimental results show SENMap enables 40 percent energy improvements for a baseline SENSIM operating in timestep asynchronous mode of operation. SENMap is designed in such a way that it facilitates mapping large spiking neural networks for future modifications as well
Active Learning for Deep Learning-Based Hemodynamic Parameter Estimation
Get PDFHemodynamic parameters such as pressure and wall shear stress play an important role in diagnosis, prognosis, and treatment planning in cardiovascular diseases. These parameters can be accurately computed using computational fluid dynamics (CFD), but CFD is computationally intensive. Hence, deep learning methods have been adopted as a surrogate to rapidly estimate CFD outcomes. A drawback of such data-driven models is the need for time-consuming reference CFD simulations for training. In this work, we introduce an active learning framework to reduce the number of CFD simulations required for the training of surrogate models, lowering the barriers to their deployment in new applications. We propose three distinct querying strategies to determine for which unlabeled samples CFD simulations should be obtained. These querying strategies are based on geometrical variance, ensemble uncertainty, and adherence to the physics governing fluid dynamics. We benchmark these methods on velocity field estimation in synthetic coronary artery bifurcations and find that they allow for substantial reductions in annotation cost. Notably, we find that our strategies reduce the number of samples required by up to 50% and make the trained models more robust to difficult cases. Our results show that active learning is a feasible strategy to increase the potential of deep learning-based CFD surrogates
Transforming Learning Environments: Asset Management, Social Innovation and Design Thinking for Educational Facilities 5.0
Get PDFEducational institutions are facing a crisis characterized by the need to address diverse learning styles and vocational aspirations, exacerbated by ongoing financial pressures. To navigate these challenges effectively, there is an urgent need to innovate educational practices and learning environments, ensuring they are adaptable and responsive to the evolving needs of students and the workforce. The adoption of the Industry 5.0 framework offers a promising solution, providing a holistic approach that emphasizes the integration of human creativity and advanced technologies to transform educational institutions into resilient, human-centric, and sustainable learning environments. In this context, this article presents a transdisciplinary methodology that integrates Asset Management (AM) with Social Innovation (SI) through Design Thinking (DT) to co-design Educational Facilities 5.0 with stakeholders. The application of the proposed approach in an AgroLab case study—a food and agricultural laboratory—demonstrates how the methodology enables the definition of an Educational Facility 5.0 and generates AM Design Knowledge to support informed decision-making in the subsequent design, implementation, and operation phases. Following DT principles—where knowledge emerges through iterative experimentation and insights from practical applications—this article also discusses the role of SI and DT in AM, the role of Large Language Models in convergent processes, and a vision for Educational Facilities 5.0
A Physiological-Model-Based Neural Network Framework for Blood Pressure Estimation from Photoplethysmography Signals
Get PDFContinuous blood pressure (BP) estimation via photoplethysmography (PPG) remains a significant challenge, particularly in providing comprehensive cardiovascular insights for hypertensive complications. This study presents a novel physiological model-based neural network (PMB-NN) framework for BP estimation from PPG signals, incorporating the identification of total peripheral resistance (TPR) and arterial compliance (AC) to enhance physiological interpretability. Preliminary experimental results, obtained from a single healthy participant under varying activity intensities, demonstrated promising accuracy, with a median standard deviation of 6.88 mmHg for systolic BP and 3.72 mmHg for diastolic BP. The median error for TPR and AC was 0.048 mmHg*s/ml and -0.521 ml/mmHg, respectively. Consistent with expectations, both estimated TPR and AC exhibited a reduction as activity intensity increased
Influence of AI Decision Support on Radiologists’ Performance and Visual Search in Screening Mammography
No full textBackground: Artificial intelligence (AI) decision support may improve radiologist performance during screening mammography interpretation, but its effect on radiologists’ visual search behavior remains unclear. Purpose: To compare radiologist performance and visual search patterns when reading screening mammograms with and without an AI decision support system. Materials and Methods: In this retrospective multireader multicase study, 12 breast screening radiologists with 4–32 years of experience (median, 12 years) from 10 institutions evaluated screening mammograms acquired between September 2016 and May 2019. Assessments were conducted unaided and with a Food and Drug Administration–approved, European Commission–marked AI decision support system, which assigns a region suspicion score from 1 to 100, with 100 indicating the highest malignancy likelihood. An eye tracker monitored readers’ eye movements. Area under the receiver operating characteristic curve (AUC), sensitivity, and specificity between unaided and AI-assisted reading were compared using multireader multicase analysis software. Reading times, breast fixation coverage (percentage breast covered by fixations within 2.5° visual angle radius) fixation time, and time to first fixation within the lesion region were compared using bootstrap resampling (n = 20 000). Results: Mammography examinations (75 with breast cancer, 75 without breast cancer) from 150 women (median age, 55 years [IQR, 50–63 years]; age range, 49–72 years) were read. The mean AUC was higher with AI support versus unaided reading (unaided, 0.93 [95% CI: 0.91, 0.96]; AI-supported, 0.97 [95% CI: 0.95, 0.98]; P < .001). There was no evidence of a difference in mean sensitivity (81.7% [735 of 900 readings] vs 87.2% [785 of 900]; P = .06), specificity (89.0% [801 of 900] vs 91.1% [820 of 900]; P = .46), or reading time (29.4 vs 30.8 seconds; P = .33). Breast fixation coverage was lower with AI support (11.1% vs 9.5% of breast area; P = .004), while fixation time in the lesion region was higher (4.4 vs 5.4 seconds; P = .006). There was no evidence of a difference in time to first fixation within the lesion region (3.4 vs 3.8 seconds; P = .13). Conclusion: Radiologists improved their breast cancer detection accuracy when reading mammography with AI support, spending more fixation time on suspicious areas and less on the rest of the breast, indicating a more efficient search.</p
Perprocedural Heparinization in Non-cardiac Arterial Procedures:The Current Practice in the Netherlands
Get PDFPurpose: Heparin is the most widely-used anticoagulant to prevent thrombo-embolic complications during non-cardiac arterial procedures (NCAP). Unfortunately, there is a lack of evidence and consequently non-uniformity in guidelines on perprocedural heparin management. Detailed insight into the current practice of antithrombotic strategies during NCAP in the Netherlands is important, aiming to identify potential optimal protocols and local differences concerning perprocedural heparinization. Materials and Methods: A comprehensive online survey was distributed electronically to vascular surgeons of every hospital in the Netherlands in which NCAP were performed. Data were collected from September 2020 to October 2021. Results: The response rate was 90% (53/59 hospitals). During NCAP, all surgeons generally administered heparin before arterial clamping. In 74% (39/54) of hospitals, a single heparin dosing protocol was used for all types of patients and vascular procedures. In 40%, there was no uniformity in heparin dosing between vascular surgeons. Depending on the procedure, a fixed bolus heparin, predominantly 5000 IU, was administered in 73% to 93%. In the remaining hospitals (7%–27%), a bodyweight-based heparin protocol was used, with an initial dose of 70 or 100 IU/kg. A minority (28%) monitored the effect of heparin in patients using the activated clotting time add (ACT) after activated clotting time. Target values varied between 180 and 250 seconds or 2 times the baseline ACT. Conclusion: This survey demonstrates considerable variability in perprocedural heparinization during NCAP in the Netherlands. Future research on heparin dosing is needed to harmonize and optimize heparin dosage protocols and contemporary guidelines during NCAP, and thereby improve vascular surgical care and patient safety. Clinical Impact: This survey demonstrated persisting intra- and inter-hospital variability in perprocedural heparinization during non-cardiac arterial procedures (NCAP) in the Netherlands. The observed variability in heparinization strategies highlights the need for high quality evidence on perprocedural anticoagulation strategies. This is needed in order to harmonize and optimize heparin dosage protocols and contemporary guidelines and thereby improve vascular surgical patient care. Based on the current results, an international survey will be conducted by the authors to gain additional insight into the antithrombotic strategies used during NCAP, aiming to harmonize anticoagulation protocols worldwide.</p