150259 research outputs found
Sort by
Deciphering aseismic deformation along submarine fault branches below the eastern Sea of Marmara (Turkey): Insights from seismicity, strainmeter, and GNSS data
No full textThe recently observed slow transients in the Sea of Marmara are important to quantify the seismic hazard and risk for the greater Istanbul metropolitan region. In this study, we analyze and characterize a slow slip event that occurred in the Eastern Sea of Marmara in 2016. To characterize the temporal history and the location of this event, we combine for the first time in this region different types of geodetic data (strainmeters and GNSS stations) and seismicity. We propose two interpretations to explain the observations: either the slow event initiated on the western part of the Armutlu fault and then propagated approximately 40 km eastward, or it initiated on the western section of the Armutlu fault, and then jumped onto a perpendicular fault after propagating ca. 15 km. We deduce these interpretations from forward modeling of the strain and displacement data. In addition, our results also suggest that this slow event triggered seismicity on a neighboring perpendicular fault
Precision cosmology with the lightest elements
No full textAbstract
In this brief review, we describe the key role played by Big Bang Nucleosynthesis (BBN) in today’s ‘Precision Cosmology’, focusing in particular on the precise determination of the primordial abundance of deuterium. We describe the development of the ideas and methods of BBN research from their beginnings more than 75 years ago to the latest developments, and conclude with a forward look to likely advances expected towards the end of the current decade.</jats:p
Viral but Vanishing: Investment Advisors, Social Media, and Regulation
No full textOver 60% of younger retail investors rely on social media platforms like StockTwits for trading insights, displacing traditional investment advice and creating a new form of financial disintermediation. This paper provides a novel contribution by examining how investment advice diffuses through financial social networks and how professionals navigate these platforms in an attempt to avoid disintermediation. Using StockTwits data, we identify users whose messages go viral, so-called effective influencers, and assess how network structure and message content shape influence. We find that professionals strategically assert dominance, being 366% more likely than regular users to post viral messages, while avoiding partisan or emotionally charged content. Despite this influence, their activity declined sharply after 2018. We show that this decline was caused by intensified SEC and FINRA oversight. Unlike historical patterns - where regulatory reforms typically prompt gradual behavioral change if there is a change at all - the StockTwits case reveals a strong and immediate regulatory impact. Overall, the findings suggest that regulation rather than platform design or user behavior caused disintermediation in investment advice, with important consequences for how retail investors access and interpret financial advice
Unveiling the lithospheric and mantle dynamic controls on Antarctic topography
No full textAccurately constraining past and future ice sheet evolution requires a better understanding of boundary conditions for ice sheet models. Two key factors - topography and heat flux - are both in part influenced by spatially and temporally variable mantle dynamics. This study uses an interdisciplinary approach to probe the present-day crust and mantle beneath Antarctica. First, observed bathymetry and topography in the oceanic and continental realms are corrected for isostatic effects to isolate the residual topographic signal as a proxy for dynamic support. In this way, a comprehensive suite of oceanic residual depth and continental residual elevation anomalies are calculated. This compilation includes the addition of 487 newly interpreted multichannel seismic lines in the Southern Ocean, and new receiver function analysis for 15 stations in the northern Transantarctic Mountains. A novel joint modelling approach, integrating high-frequency receiver functions and P-wave coda autocorrelation, is used to constrain ice and subglacial properties. Subsequently, inversion of low-frequency receiver functions reveals crustal velocity structure, refining estimates of bulk crustal thickness and density, as required for residual elevation calculations. All resulting residual depth and elevation measurements are integrated into previously published global databases to present a smoothed global spherical harmonic representation of the data. While high spherical harmonic degrees (l = 40) effectively capture oceanic swells in mantle support, their significance to continental residual elevation remains limited by data uncertainty. Nevertheless, spatial patterns of inferred dynamic topography in the continental and ocean realms are consistent with upper mantle (<300km) shear wave velocity anomalies, long-wavelength free-air gravity anomalies, regions of thinned lithosphere, and Neogene intraplate basaltic volcanism. Modelling of rare earth element concentrations from these volcanic provinces defines melt fraction as a function of depth, allowing for both mantle potential temperature and lithospheric thickness to be constrained. Results are consistent with geophysical observations and reinforce understanding that regions of upwelling mantle act to thin the lithosphere and elevate topography. Steeper geothermal gradients associated with regions of plate thinning and volcanism have significant implications for the delivery of heat to the base of the Antarctic Ice Sheet. The high average elevation of East Antarctica, however, is attributable to buoyancy induced by lithospheric depletion, acting to thermally insulate the cratonic region from mantle processes. Collectively, these observations highlight the role of mantle-lithosphere interactions in Antarctica and the subsequent importance of accounting for how convective processes have shaped topography through time
Randomised Badger Culling Trial—no effects of widespread badger culling on tuberculosis in cattle: comment on Mills, Woodroffe and Donnelly (2024a, 2024b)
No full textSubstantial problems remain with the Mills et al. (Mills CL, Woodroffe R, Donnelly CA. 2024 R. Soc. Open Sci.11, 240385. (doi:10.1098/rsos.240385) and Mills CL, Woodroffe R, Donnelly CA. 2024 R. Soc. Open Sci.11, 240386. (doi:10.1098/rsos.240386)) re-evaluations of the Randomised Badger Culling Trial (RBCT) data. These include: (i) analysis of counts rather than rates; (ii) overfitting; (iii) modelling of ‘confirmed’ herd breakdowns rather than total herd breakdowns; and (iv) errors in the Bayesian analysis of the data. More plausible approaches to RBCT data analysis, including analysis of total breakdown incidence, strongly suggest that there is no effect of proactive badger culling on the herd incidence of bovine tuberculosis, either within cull areas or in neighbouring areas
Peptide cages: bioinspired supramolecular architectures for next-generation applications
No full textThis review analyses peptide-based supramolecular cages, their design and applications.</jats:p
Exploring medical co-design in the AI era: A scoping review
No full textObjective This study explores the current development and findings of medical co-design (MCD) in the era of artificial intelligence (AI). As design objects evolve from products addressing single-function needs to complex systems catering to diverse stakeholders' values, medical designers face the challenges of overcoming high knowledge barriers and coordinating multiple interests. MCD is regarded as an effective approach to addressing these challenges. Methods This study adopts a scoping review approach to analyse the progress and limitations of existing research in the field of MCD, with a particular focus on overcoming knowledge barriers. Additionally, the study examines the integration of AI technologies, particularly the application of large language models (LLMs), exploring their evolving roles in co-design and their impact on participants, providing new insights into human–AI co-design. Results The study reveals that traditional human-centred co-design has gradually evolved into human–AI co-design and even AI–AI co-design. This paper presents a theoretical framework for multi-agent collaboration, preliminarily identifying the interaction patterns between humans and AI agents during the design process and their potential impact on design outcomes. Conclusion This study offers a theoretical foundation and practical guidance for future AI-assisted MCD practices. The findings contribute to a better understanding of the intrinsic mechanisms of human–AI co-design and provide insights for optimising knowledge integration and innovation processes in MCD
Closed-Loop Therapy and Sleep in Young People Newly Diagnosed With Type 1 Diabetes and Their Parents.
No full textBACKGROUND: A diagnosis of type 1 diabetes in a young person can create vulnerability for sleep. Historically it has been rare for young people to be offered a closed-loop system soon after diagnosis meaning that studies examining sleep under these circumstances in comparison with standard treatment have not been possible. In this study, we examine sleep in young people (and their parents) who were provided with hybrid closed-loop therapy at diagnosis of type 1 diabetes versus those who receive standard treatment over a 2-year period. METHODS: The sample comprised 97 participants (mean age = 12.0 years; SD = 1.7) from a multicenter, open-label, randomized, parallel trial, where young people were randomized to either hybrid closed-loop insulin delivery or standard care at diagnosis. Sleep was measured using actigraphy and the Pittsburgh Sleep Quality Index (PSQI) in the young people, and using the PSQI in parents. RESULTS: Sleep in young people using hybrid closed-loop insulin delivery did not differ significantly compared with those receiving standard care (although there were nonsignificant trends for better sleep in the closed-loop group for 4 of the 5 sleep actigraphy measures and PSQI). Similarly, there were nonsignificant differences for sleep between the groups at 24 months (with mixed direction of effects). CONCLUSIONS: This study assessed for the first time sleep in young people using a closed-loop system soon after diagnosis. Although sleep was not significantly different for young people using closed-loop insulin delivery as compared with those receiving standard care, the direction of effects of the nonsignificant results indicates a possible tendency for better sleep quality in the hybrid closed-loop insulin delivery group at the beginning of the treatment
Categorical models of second-order abstract syntax
No full textMathematics and computer science increasingly rely on proof assistants to verify reasoning and ensure program correctness. Yet a persistent obstacle in the formalisation of programming languages and calculi is the treatment of variables and the associated operations of α-renaming and capture-avoiding substitution. Despite many proposed approaches, none match the flexibility and clarity of informal reasoning on paper. As a result, formalising languages in proof assistants often demands navigating a cumbersome layer of syntactic metatheory before any real benefits of mechanisation can be realised.
In parallel, mathematical frameworks offer powerful, reusable tools for working with syntax – such as type-preserving simultaneous substitution and compositional semantics via initial algebras. However, their practical impact on formal verification has been limited, due to their categorical sophistication and the challenges of encoding them in dependently-typed settings.
This thesis bridges this gap by formally relating the method of intrinsically typed encodings to the presheaf approach for languages with binding. We introduce the familial model of second-order abstract syntax as a categorical foundation for intrinsic typing, and show its equivalence to the presheaf model both syntactically and semantically. This places many ad hoc practices on solid mathematical footing, opening up principled paths for abstraction and extension. Along the way, we develop general tools for weakened monoidal structures and functorial models of syntax, framed by adjoint modalities – laying the groundwork for scalable and mechanised reasoning about syntax
Composite grey matter fingerprints for genetic frontotemporal dementia.
No full textBACKGROUND: Brain structural changes in frontotemporal dementia (FTD) can occur decades before symptom onset. Precise characterisation of grey matter changes is necessary for developing models of biomarker progression, while better understanding the trajectory of the pathology is invaluable for prognosis and detecting treatment effects as we enter the era of clinical trials. METHODS: Cortical and subcortical grey matter volume and thickness from structural MRI were assessed in a large cohort of 892 participants including presymptomatic and symptomatic carriers of mutations within the three main genetic causes of FTD (C9 open reading-frame 72 (C9orf72), progranulin (GRN) and microtubule-associated protein tau (MAPT)) compared with mutation-negative relatives (controls). We compared the distribution of grey matter changes of each metric at different stages of the disease cross sectionally. We aimed to identify grey matter composites for each genetic group which would show the earliest changes and which separated presymptomatic carriers from controls. RESULTS: While C9orf72 mutation carriers showed widespread presymptomatic grey matter changes, MAPT and particularly GRN mutation carriers showed changes more proximally to symptom onset. Our composite grey matter signatures, which discriminate asymptomatic/prodromal carriers from controls with high to very high areas under the curve, involved bilateral thalami volumes, precuneus and postcentral thickness in C9orf72; left caudal middle frontal thickness, frontal pole and pars orbitalis volumes in GRN; right temporal pole volume and left insula thickness in MAPT mutation carriers. CONCLUSION: We propose the use of cortical thickness and volume measurements combined from multiple regions into a composite region of interest for each FTD genetic group to identify the earliest changes and track disease progression. Our quasi-longitudinal design illustrates that these regions continue to evolve throughout the symptomatic stages. Investigating how our selected composites progress and validating these in longitudinal samples will be invaluable for future clinical trials