435032 research outputs found
Sort by
Revisiting the purpose of the thalamus:anatomically sub-cortical but functionally supra-cortical?
Get PDFThe largely accepted view of the brain function is corticocentric: the cerebral cortex monitors the world to provide optimal behavioural decisions given current external and internal demands. However, this overlooks a crucial fact: the cerebral cortex has very little direct input from the outside world, and hence is not in a position to monitor it. In fact, the cortex receives the majority of its information from the thalamus. As such, the cortex monitors the thalamus, while the thalamus monitors the external environment. Interactions between the two are controlled by inhibition from the thalamic reticular nucleus (TRN), which dynamically modifies the information transmitted by the thalamus to the cortex and vice versa. The thalamus is an evolutionarily conserved structure, present in all vertebrates. While it is increasingly clear that the thalamus is not a simple sensory relay, the exact purpose of this structure and of the thalamocortical architecture which separates the computational power of the cerebral cortex from direct access to information about the external world is not clear. Theoretical considerations regarding self-monitoring of brain function and observations of the behavioural impact of resective surgery elevate the thalamus in control terms from anatomically sub-cortical to functionally supra-cortical. They suggest that the final output of the system is determined, dynamically, at the thalamic level, driven by the integration of the cortical and sensory inputs with the current thalamic state. Investigating this idea to understand the role of the thalamus will require a concerted effort across electrophysiology, neuroimaging and computational approaches.</p
Beyond elevation:Unravelling land cover–elevation interactions in governing soil moisture variability in the lower Himalayas
Get PDFUnderstanding the spatiotemporal variability of soil moisture (SM) in topographically complex and data-scarce regions, such as the Himalayas, remains a major challenge due to the interactive influence of elevation, land cover heterogeneity, and anthropogenic activities. This study presents an extensive 57-week soil moisture monitoring campaign conducted at 104 systematically gridded locations in the Suketi watershed, Himachal Pradesh, India. Soil moisture samples were classified into three land cover types: agricultural lands, forests, and grasslands, and further stratified into five elevation zones (739–2914 m) using a quantile-based classification method to accurately capture altitudinal variability. Statistical methods, including Student’s t-distribution, Spearman rank correlation, analysis of variance, Tukey’s post-hoc test, and partial dependence plots, were employed to quantify the interactive controls of SM variability. The findings underscore that elevation alone inadequately explains SM variability, however, land cover-specific attributes such as vegetation cover, soil texture, and anthropogenic practices exert significant influence. Grasslands consistently exhibited higher moisture retention (∼18.05 %) and lower spatial variability across elevations, making them ideal buffer zones for hydrological planning. Forests exhibited moderate SM levels (∼16.69 %) and greater temporal stability, while agricultural areas showed pronounced variability (standard deviation ∼5.20 %) and significant sensitivity to elevation (ρ = –0.36, p < 0.01), indicating the need for denser sampling strategies in such landscapes. The study also introduces a statistically robust framework for determining optimal sampling locations under varying precision thresholds, offering valuable guidance for sensor network optimization. The findings provide critical insights for enhancing hydrological modeling, satellite SM product validation, and precision water resource management in mountainous environments
Measurement of the lithographic point-spread function of a focused helium ion beam in negative-tone PMMA and fullerene resists on ultrathin membranes
Get PDFHelium-ion-beam lithography has many advantages relevant to the fabrication of dense arrays of nanostructures such as a small probe size, large depth of field, and reduced proximity effect. Here, we calculate and measure the lithographic point-spread functions (PSFs) of 30 keV He+ ions in negative-tone polymethyl methacrylate (PMMA) and a fullerene-derivative resist on ultrathin silicon nitride membranes and compare the results to similar work by Manfrinato et al. (2017) measuring the PSF of 200 keV electrons in PMMA using an aberration-corrected scanning transmission electron microscope (STEM). PSFs were calculated using the method reported previously by Winston et al. (2012). Our results show that both the He+ ion/PMMA and He+ ion/fullerene-derivative resist PSFs decay more rapidly with distance, r, from the point of incidence of the beam than the corresponding aberration-corrected electron beam/PMMA PSF. In fact, the He+ ion PSFs decay approximately with r−4 while the aberration-corrected EBL PSF decays approximately with r−2. This result implies that the lateral area exposed by the focused beam increases more rapidly with dose for e− beams than He+ beams. Effectively, this should result in reduced proximity effect in helium-ion-beam lithography. This work provides further evidence that HIBL offers distinct advantages over EBL for high-resolution and high-density patterning as well as highlighting some benefits of the fullerene-derivative resist over PMMA.</p
A coupled FEM–SBM methodology for dynamic interaction of multiple structures and soil
No full textIn densely populated urban areas, there is a growing trend in constructing complex structural systems that include underground structures located beneath clusters of aboveground buildings. The dynamic interaction between these underground and aboveground structures, mediated by the surrounding soil, is known as structure-soil–structure interaction (SSSI). SSSI is a topic whose effects pose significant challenges to the design and analysis of such complex systems. In this paper, we propose a novel numerical methodology for addressing longitudinally invariant multi-structure-soil interaction problems in elastodynamics. The proposed approach combines the Finite Element Method (FEM) for modelling structural components with the Singular Boundary Method (SBM) for simulating wave propagation in the soil and capturing inter-structural coupling effects, all formulated in the wavenumber-frequency domain. The synergy of FEM, well-suited to complex geometries, and the computational simplicity and efficiency of SBM yields a robust and accurate framework for solving SSSI problems. The framework features a strongly coupled formulation between structures and soil, enhancing both accuracy and ease of implementation. The accuracy of the method is assessed through several benchmark studies involving cylindrical shells and cylindrical solids, while its practical applicability is demonstrated via real-scale numerical examples, with relative errors typically below 2%. Furthermore, the computational efficiency of the proposed methodology is compared with traditional hybrid approaches, in which both the structures and the surrounding soil are modelled using FEM, with the remaining soil represented via the Method of Fundamental Solutions (MFS) or Boundary Element Method (BEM). On average, the proposed approach achieves a computational performance nearly twenty times faster than that of the reference solutions. The results underscore the advantages of the proposed framework in terms of modelling simplicity, numerical efficiency, accuracy and robustness, and show that the method is scalable and capable of evaluating interactions among an arbitrary number of structures
LungDetectNet: a multi-task deep learning framework with enhanced detection and descriptive capabilities
Get PDFArtificial intelligence has significantly transformed medical image analysis, particularly in the early diagnosis of lung cancer from computed tomography (CT) scans. A key step in this diagnostic process is the accurate identification of lung nodules, which are primary indicators of potential malignancy, yet this identification task remains challenging due to their small size and subtle features. While efficient 3D object detection frameworks like MedYOLO offer a promising approach, their optimal architecture for this specific task is not well-established. To address this, we introduce LungDetectNet, a framework that advances the MedYOLO approach for 3D lung nodule detection. This advancement was achieved through a systematic investigation of the YOLO backbone’s evolution, which demonstrated a clear performance improvement corresponding with the integration of more advanced backbone architectural designs. In addition to detection, regression heads are integrated into the framework to predict seven descriptive attributes of each detected nodule: Subtlety, Sphericity, Margin, Lobulation, Spiculation, Texture, and Malignancy. On our newly established, challenging benchmark from the LIDC-IDRI dataset, our final model achieves a Mean Average Precision (MAP) of 0.793, with a precision of 0.896 and a recall of 0.664. For the multi-task regression objective, the model also shows strong performance, achieving an average Mean Absolute Error (MAE) of 0.516. These results demonstrate that LungDetectNet has the potential to enhance early-stage lung nodule detection and provide detailed diagnostic insights, therefore supporting clinical decision-making and improving patient care.<br/
A comparative investigation of the impact of digitalisation and work-from-home policy on firm performance:MNEs vs international SMEs
Get PDFThe COVID-19 pandemic accelerated the adoption of work-from-home (WFH) practices, raising important questions about their long-term implications for organisational performance. This issue is particularly salient for multinational enterprises (MNEs) and international small and medium-sized enterprises (SMEs), where digitalisation has significantly reshaped work arrangements. This study evaluates the advantages and disadvantages of WFH and their differential impacts on MNEs and international SMEs. A comparative analysis was conducted using expert pairwise judgements, assessed through linguistic terms with weakened hedges (LTWHs) within the Best–Worst Method (BWM) framework. The LTWHs approach enables experts to articulate nuanced and flexible preferences, extending traditional linguistic scales by softening the strength of evaluations. This makes it particularly suited for capturing subjective assessments of WFH impacts under conditions of uncertainty. The findings indicate substantial variation in WFH adoption, with private sector organisations demonstrating approximately 50 % greater willingness to adopt WFH compared to public authorities. The analysis further highlights the distinct advantages and disadvantages that shape the performance outcomes of MNEs and international SMEs. By introducing a novel hesitant fuzzy linguistic preference approach, this study develops a comprehensive framework for assessing the organisational consequences of WFH. The results offer valuable insights for managers and policymakers seeking to strike a balance between flexibility, productivity, and resilience in the design of post-pandemic work strategies
Hypercontractivity beyond Nelson's time and its applications to Blaschke--Santaló inequality and inverse Santaló inequality
No full textTransferable Model-Based Reinforcement Learning for Vehicular Platoon Control
Get PDFThe learning efficiency remains a critical impediment to the practical application of connected and automated vehicles (CAVs). This paper proposes a transferable model-based reinforcement learning (TMBRL) strategy to enhance the sample efficiency and learning rate of CAVs. Specifically, a surrogate policy model is established by capturing state transitions between the actual environment and the vehicle within traffic scenarios. Then, a model-based reinforcement learning (MBRL) approach is established utilizing a surrogate model and a soft actor-critic algorithm. To improve the learning efficiency of platoon control algorithm, a transfer learning method is implemented to MBRL framework. Specifically, the trained surrogate model of vehicles in the source domain is transferred to vehicles in the target domain, and the latter just should update the surrogate model in terms of the individual dynamic characteristics and tasks. Finally, a platoon experiment platform with Prescan software is conducted. The experimental evaluation demonstrates that the TMBRL strategy significantly outperforms conventional reinforcement learning approaches, achieving higher average cumulative reward of 47 and demonstrating a 16% improvement in training success rate. Comparative analysis further reveals that the proposed TMBRL strategy exhibits superior robustness in platoon tracking tasks, maintaining enhanced trajectory tracking precision and stability under dynamic environmental conditions
Digital transformation through innovation:The human-AI decision spectrum
Get PDFThe rapidly changing and increasingly complex processes enabled by artificial intelligence (AI) applications challenge the conventional concepts of innovation. In contrast to a general perception that AI adoption can augment innovation output, managers still lack empirical guidance on how to structure innovation processes with human-AI interaction across time and space. Drawing on observations from case studies in the aerospace, heavy engineering, information technology, and pharmaceutical sectors, this paper presents the development of a conceptual model for digital innovation to represent (i) Learning Processes (LPs) focusing on knowledge creation and knowledge reuse and (ii) Product Development Processes (PDPs) leading to radical and incremental changes. The conceptual model is inductively developed based on a theory building approach using multiple case studies. A set of transformative characteristics centralized on Originality, Reliability, Transferability, and Adaptability (ORTA) are identified to guide decision-making along multi-stage and cross-layer innovation processes involving cyclical handoffs between humans and machine agents. These ORTA characteristics form a base for strategic decision-making along the Human and AI decision spectrum suited to prepare companies for survival and prosperity in their journeys of digital transformation
Methodology of the Impact360 Traffic Tool:Working Paper
Get PDFThe traffic tool is part of an Impact360 tool that assesses the economic, environmental and social impacts of street works and road maintenance. The Impact360 tool enables optioneering on how to carry out the works, and can be found at: https://nzt-carbon-calculator-web-app-service-qa.azurewebsites.net. The traffic tool methodology is by the University of Birmingham, and the user interface and coding is by EA Technology. It is funded by Transport for London and Kent County Council
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇