36270 research outputs found
Sort by
MICCAI STS 2024 Challenge: Semi-Supervised Instance-Level Tooth Segmentation in Panoramic X-ray and CBCT Images
No full textOrthopantomogram (OPGs) and Cone-Beam Computed Tomography (CBCT) are vital for dentistry, but creating large datasets for automated tooth segmentation is hindered by the labor-intensive process of manual instance-level annotation. This research aimed to benchmark and advance semi-supervised learning (SSL) as a solution for this data scarcity problem. We organized the 2nd Semi-supervised Teeth Segmentation (STS 2024) Challenge at MICCAI 2024. We provided a large-scale dataset comprising over 90,000 2D images and 3D axial slices, which includes 2380 OPG images and 330 CBCT scans, all featuring detailed instance-level FDI annotations on part of the data. The challenge attracted 114 (OPG) and 106 (CBCT) registered teams. To ensure algorithmic excellence and full transparency, we rigorously evaluated the valid, open-source submissions from the top 10 (OPG) and top 5 (CBCT) teams, respectively. All successful submissions were deep learning-based SSL methods. The winning semi-supervised models demonstrated impressive performance gains over a fully-supervised nnU-Net baseline trained only on the labeled data. For the 2D OPG track, the top method improved the Instance Affinity (IA) score by over 44 percentage points. For the 3D CBCT track, the winning approach boosted the Instance Dice score by 61 percentage points. This challenge demonstrates the potential benefit benefit of SSL for complex, instance-level medical image segmentation tasks where labeled data is scarce. The most effective approaches consistently leveraged hybrid semi-supervised frameworks that combined knowledge from foundational models like SAM with multi-stage, coarse-to-fine refinement pipelines. Both the challenge dataset and the participants’ submitted code have been made publicly available on GitHub (https://github.com/ricoleehduu/STS-Challenge-2024), ensuring transparency and reproducibility.</p
Can GRB 250702B be explained as the tidal disruption of a white dwarf by an intermediate mass black hole? Yes
No full textGRB 250702B is a unique astrophysical transient characterised by its nature as a repeating gamma-ray trigger. Its properties include possible periodicity in its gamma-ray light curve, an X-ray counterpart that rose prior to the gamma-ray outbursts and faded quickly, and radio and infrared counterparts. These features are difficult to reconcile with most models of high energy transients but we show that they are compatible with a white dwarf bound to an intermediate mass black hole that is tidally stripped over multiple pericentre passages before being fully disrupted. In this model, accretion onto the black hole powers a mildly relativistic jet that produces the X-rays through internal processes and the infrared and radio counterparts through thermal emission and external shocks respectively but is unable to produce the gamma-ray emission on its own. We find that chaotic debris streams from the multiple stripping episodes can collide with a period roughly the same as the orbital period of the star. These shocks produce X-ray photons that are upscattered by the jet to produce the observed MeV gamma-ray emission. Future analysis of the jet properties will allow us to place firmer constraints on our model.</p
Tree ring isotopes reveal an intensification of the hydrological cycle in the Amazon
No full textOver recent decades the Amazon region has been exposed to large-scale land-use changes and global warming. How these changes affect Amazonia’s hydrological cycle remains unclear as meteorological data are scarce. We use tree ring oxygen isotope records to confirm that the Amazon hydrological cycle has intensified since 1980. Diverging isotopic trends from terra firme and floodplain trees from distinct sites (approximately 1000 km apart) in Western Amazon indicate rainfall amounts increased during the wet season and decreased during the dry season at large-scale. Using the Rayleigh distillation model, we estimate that wet season rainfall increased by 15–22%, and dry season rainfall decreased by 8–13%. These diverging trends provide evidence, independent from existing climate records, that the seasonality of the hydrological cycle in the Amazon is increasing. Continuation of the observed trends will have a pervasive impact on Amazon forests and floodplain ecosystems, and strongly affect the livelihoods of the regional riverine communities.</p
SLENet: A Guidance-Enhanced Network for Underwater Camouflaged Object Detection
No full textUnderwater Camouflaged Object Detection (UCOD) aims to identify objects that blend seamlessly into underwater environments. This task is critically important to marine ecology. However, it remains largely underexplored and accurate identification is severely hindered by optical distortions, water turbidity, and the complex traits of marine organisms. To address these challenges, we introduce the UCOD task and present DeepCamo, a benchmark dataset designed for this domain. We also propose Semantic Localization and Enhancement Network (SLENet), a novel framework for UCOD. We first benchmark state-of-the-art COD models on DeepCamo to reveal key issues, upon which SLENet is built. In particular, we incorporate Gamma-Asymmetric Enhancement (GAE) module and a Localization Guidance Branch (LGB) to enhance multi-scale feature representation while generating a location map enriched with global semantic information. This map guides the Multi-Scale Supervised Decoder (MSSD) to produce more accurate predictions. Experiments on our DeepCamo dataset and three benchmark COD datasets confirm SLENet’s superior performance over SOTA methods, and underscore its high generality for the broader COD task.</p
Mitochondrial adaptations in skeletal muscle following incretin-based therapies: in vitro
No full textNo description supplied</p
The environmental and sedimentary effects of a major asteroid impact-event: the Stac Fada (Stoer) Group of NW Scotland
No full textThe late Proterozoic Stac Fada Member, NW Scotland, provides 50 km of excellent exposure of a suevite sheet deposited from an impact-triggered impactoclastic density current, enclosed in sediments that reveal the landscape, environments and sedimentation events before, during and after a terrestrial hypervelocity impact.Paleocurrent data and relative sediment accumulation rates indicate a semi-arid alluvial basin within a north-south continental rift, with highlands to the east. Facies associations indicate alluvial fans, braided rivers, floodplains with ephemeral ponds and wind-blown dune fields. One facies displays textures that may record microbial mat grounds. Unusual poorly sorted sandy muds may have formed by mixing of sand into mud during repeated wetting and drying when aggradation was very slow. This process may be a feature of pre-vegetated subaerial environments and could prove useful in the study of Precambrian palaeosols.Pre-impact sediments provide information on highland basins, lower hillslopes, basin margins, ponded floodplains, braided rivers and windblown dune field settings. The suevite drapes the topography at these sites, ranging in thickness from 2 m in highlands to 11 m in the alluvial basin. The influx of loose suevite caused catastrophic sedimentation, as occurs after large explosive volcanic eruptions, involving a waning succession of deposit-derived impactoclastic density currents, debris flows, hyperconcentrated floods and clear water floods. After a short period (days to weeks) the environment gradually stabilised. Above the suevite, sandstones contain impact-melt fragments and record fluvial sedimentary reworking. However, nickel concentrations (a proxy for proportion of impact-derived material) remain above pre-impact levels much higher in the succession (~13 m) than any visible field evidence for suevite reworking. Overall, post-impact disruption to sedimentation is limited compared with that which follows catastrophic explosive volcanic eruptions, perhaps because the modern exposures are not sited down current of, or in the same catchment as, the (unexposed) impact crater.</p
On implementation of nonlocal terms in models in ecology and evolution. Comment on 'Nonlocal models in biology and life sciences: Sources, developments, and applications' by S. Pal & R. Melnick.
No full textMathematical modellling and computer simulations have been successfully applied to quantify and predict functioning of biological systems across time and space scales. Very first models in mathematical biology were based on ‘local’ (i.e. non-integral) terms in differential equations or/and discrete time maps, notable examples being: single species logistic equation, Lotka-Volterra predator-prey model, Kermack - McKendrick SIR disease model, Fisher-KPP reaction-diffusion equation, and others. However, it was shortly recognized that modelling a realistic biological object would require implementation of nonlocal terms of various nature. In ecology, the need for using nonlocal terms is often due to population structuring (in terms of age, size, life history traits, behaviour, etc), or due to the difference of characteristic time and space scales governing various processes, as reproduction, growth, movement, competition, etc. For example, for most aquatic food web models with the vertical spatial resolution (hundreds of models of this type exist in the literature, including advanced eco-climate models), one needs to take into account the self-shading of the sunlight by phytoplankton, which is described by a nonlocal term integrating the phytoplankton density from the water surface to the given depth [1], [2]. Therefore, nonlocal terms describe cumulative effects of processes over the time, physical space, as well as the space of life history traits.</p
Change in body mass index and disease burden among people with obesity and multiple long-term conditions.
No full textAIMS: Obesity is a risk factor for multiple long-term conditions (MLTCs/multimorbidity). However, the impact of weight loss in people with MLTCs is unclear. We investigated the association between body mass index (BMI) change and the development of obesity-related complications (ORCs), as well as clinical and economic outcomes in individuals with obesity and MLTCs. MATERIALS AND METHODS: This cohort study included adults aged ≤70 years with BMI ≥30 kg/m2 and ≥2 ORCs. BMI was recorded during Years 1 and 4 of the baseline period. BMI change was categorised as increases or decreases of 3%-7%, 7%-15% and 15%-30%. Data were analysed using Cox regression (hazard ratios [HRs] for mental-health conditions, hospitalisation and mortality) and Ghosh-Lin models (risk of ≥1 new ORC and clinical consultations). RESULTS: A total of 618 426 individuals were included. The mean number of ORCs at index (Year 4) was 3.8 (standard deviation 1.5). HRs (95% confidence intervals) for incident ORCs were 0.96 (0.94-0.98), 0.98 (0.96-0.99) and 0.98 (0.97-0.99) for the highest to lowest BMI reductions, respectively; BMI increases were associated with HRs >1.00. Risk ratios for consultations were 0.99 (0.98-1.00), 0.99 (0.98-1.00) and 1.00 (0.99-1.00). BMI reductions were linked to lower polypharmacy rates. Both BMI decrease and increase were associated with higher HRs for mental-health conditions, hospitalisation and mortality versus stable BMI. CONCLUSIONS: Weight loss in individuals with obesity and MLTCs was linked to both favourable and adverse outcomes, highlighting the importance of personalised treatment approaches that consider outcomes beyond weight loss.</p
Seismological Characterization of Northern Hikurangi Margin Slow Slip Regions Associated With Normal Faults, Seamounts, and Seeps
No full textAt the northern Hikurangi margin, Aotearoa New Zealand, slow slip events (SSEs) recur every 6–24 months to 30 km depth. Although shallow SSEs (0–10 km) are well‐studied offshore, the deeper portion (10–30 km) remains poorly understood, limiting insight into SSE initiation. Here we investigate this deeper region and examine relationships between newly resolved SSEs and seismicity. Using time‐dependent inversions, we resolve two small SSEs ( 6.2 and 6.4), including one that extends from 15 to 30 km depth. Using data from a dense onshore seismograph network deployed directly above this deeper portion from December 2017 to October 2018, we construct a catalog of 3,071 high‐quality earthquakes with hypocentral uncertainties 5 km, located using a 3‐D velocity model and a new 1‐D model. Earthquake magnitudes range from −0.84 to 4.40, with a completeness magnitude of 1.7 and a b‐value of 1.06. Focal mechanisms reveal numerous normal‐faulting earthquakes, including some within the slab mantle. Vertically‐aligned seismicity and normal‐faulting earthquakes outline pathways linking the slab mantle to surface seeps of mantle‐derived fluids. We infer that normal faults form due to slab bending and localized uplift of subducting seamounts, which roughen the plate interface, damage the upper plate, and promote fluid migration. Landward of 100 km from the trench, both surface seeps and normal‐faulting mechanisms cease, coinciding with the downdip limit of shallow SSEs. Together, these results suggest that the Hikurangi margin's rough subducting plate interface exerts strong control on forearc dewatering and SSE genesis.</p
Registration of Stillbirth After Early Fetal Death: A National Population-Based Analysis
No full textStillbirth rates are a vital indicator of the quality of care in pregnancy and childbirth and of woman's health [1]. Accurate information on stillbirths is required for monitoring purposes, often utilising civil registrations of stillbirths. However, there are significant methodological challenges in ensuring robust stillbirth statistics, which are essential for accurate surveillance, policy development, and international comparisons [2]. Recommendations have been developed to overcome challenges relating to the inclusion of deaths following termination of pregnancy [3] and inconsistencies in the determination of signs of life [4].In England and Wales, we have identified an additional challenge arising from the use of stillbirth registration data provided by the Office for National Statistics (ONS) for monitoring mortality. This relates to scenarios where the birth occurs after the registration threshold of 24 weeks’ gestation, but fetal death occurred in utero before this time. The Royal College of Obstetricians and Gynaecologists developed Good Practice Paper No. 4 [5] specifically recommending that a death should not be registered as a stillbirth where fetal death occurred in utero before 24 weeks' gestation, regardless of gestational age at birth. However, anecdotal evidence has suggested that this guidance is not routinely followed in practice, which could lead to inconsistencies in official statistics, inequities for parents, and misclassification of perinatal deaths. This issue is of particular concern in the case of multiple births, where the interval between fetal death and birth is often more prolonged, with birth occurring when one or more surviving babies are born.To explore these potential registration inconsistencies and examine their impact on stillbirth registration practice, we utilised national surveillance data from MBRRACE-UK (Mothers and Babies: Reducing Risk through Audits and Confidential Enquiries across the UK) [6], which includes information on both the gestation when the in utero death was confirmed and the gestation at birth. Here, we report the consistency in the implementation of RCOG GPP No. 4 regarding registration of deaths where in utero death occurs before 24 weeks, but birth occurs after the 24-week stillbirth registration threshold.</p