Explore the vast range of titles available.

The microstructures, major- and trace-element compositions of minerals and electron backscattered diffraction (EBSD) maps of high- and low-Cr# [spinel Cr# = Cr 3+ /(Cr 3+  + Al 3+ )] chromitites and dunites from the Zedang ophiolite in the Yarlung Zangbo Suture (South Tibet) have been used to reveal their genesis and the related geodynamic processes in the Neo-Tethyan Ocean. The high-Cr# (0.77–0.80) chromitites (with or without diopside exsolution) have chromite compositions consistent with initial crystallization by interaction between boninitic magmas, harzburgite and reaction-produced magmas in a shallow, mature mantle wedge. Some high-Cr# chromitites show crystal-plastic deformation and grain growth on previous chromite relics that have exsolved needles of diopside. These features are similar to those of the Luobusa high-Cr# chromitites, possibly recycled from the deep upper mantle in a mature subduction system. In contrast, mineralogical, chemical and EBSD features of the Zedang low-Cr# (0.49–0.67) chromitites and dunites and the silicate inclusions in chromite indicate that they formed by rapid interaction between forearc basaltic magmas (MORB-like but with rare subduction input) and the Zedang harzburgites in a dynamically extended, incipient forearc lithosphere. The evidence implies that the high-Cr# chromitites were produced or emplaced in an earlier mature arc (possibly Jurassic), while the low-Cr# associations formed in an incipient forearc during the initiation of a new episode of Neo-Tethyan subduction at ~130–120 Ma. This two-episode subduction model can provide a new explanation for the coexistence of high- and low-Cr# chromitites in the same volume of ophiolitic mantle.

The intertidal bivalve Katelysia rhytiphora, endemic to south Australia and Tasmania, is used here for pulsed Sr-labelling experiments in aquaculture experiments to visualize shell growth at the micro- to nanoscale. The ventral margin area of the outer shell layer composed of (i) an outermost outer shell layer (oOSL) with compound composite prismatic architecture with three hierarchical orders of prisms and (ii) an innermost outer shell layer (iOSL) with crossed-acicular architecture consisting of intersecting lamellae bundles. All structural orders in both layers are enveloped by an organic sheath and the smallest mineralized units are nano-granules. Electron backscatter diffraction reveals a strong preferred orientation of the aragonite c axes perpendicular to the growth layers, while the a and b axes are scattered within a plane normal to the local growth direction and >46 % twin grain boundaries are detected. The Young's modulus shows a girdle-like maximum of elastically stiffer orientations for the shell following the inner shell surface. For 6 d, the bivalves were subjected twice to seawater with an increased Sr concentration of 18× mean ocean water by dissolving 144 µg g−1 Sr (159.88 Sr∕Ca mmol ∕ mol) in seawater. The pulse labelling intervals in the shell are 17× (oOSL) and 12× (iOSL) enriched in Sr relative to the Sr-spiked seawater. All architectural units in the shell are transected by the Sr label, demonstrating shell growth to progress homogeneously instead of forming one individual architectural unit after the other. Distribution coefficients, DSr ∕ Ca, for labelled and unlabelled shells are similar to shell proportions formed in the wild (0.12 to 0.15). All DSr ∕ Ca values are lower than values for equilibrium partitioning of Sr in synthetic aragonite.

To assess the quality of the arterial input function (AIF) reconstructed using a dedicated pre-bolus low-dose contrast material injection imaged with a high temporal resolution and the resulting estimated liver perfusion parameters. In this IRB-approved prospective study, 24 DCE-MRI examinations were performed in 21 patients with liver disease (M/F 17/4, mean age 56 y). The examination consisted of 1.3 mL and 0.05 mmol/kg of gadobenate dimeglumine for pre-bolus and main bolus acquisitions, respectively. The concentration-curve of the abdominal aorta in the pre-bolus acquisition was used to reconstruct the AIF. AIF quality and shape parameters obtained with pre-bolus and main bolus acquisitions and the resulting estimated hepatic perfusion parameters obtained with a dual-input single compartment model were compared between the 2 methods. Test-retest reproducibility of perfusion parameters were assessed in three patients. The quality of the pre-bolus AIF curve was significantly better than that of main bolus AIF. Shape parameters peak concentration, area under the time activity curve of gadolinium contrast at 60 s and upslope of pre-bolus AIF were all significantly higher, while full width at half maximum was significantly lower than shape parameters of main bolus AIF. Improved liver perfusion parameter reproducibility was observed using pre-bolus acquisition [coefficient of variation (CV) of 4.2%-38.7% for pre-bolus vs. 12.1-71.4% for main bolus] with the exception of distribution volume (CV of 23.6% for pre-bolus vs. 15.8% for main bolus). The CVs between pre-bolus and main bolus for the perfusion parameters were lower than 14%. The AIF reconstructed with pre-bolus low dose contrast injection displays better quality and shape parameters and enables improved liver perfusion parameter reproducibility, although the resulting liver perfusion parameters demonstrated no clinically significant differences between pre-bolus and main bolus acquisitions.

Whether the route of early feeding affects outcomes of patients with severe critical illnesses is controversial. We hypothesised that outcomes were better with early first-line enteral nutrition than with early first-line parenteral nutrition. In this randomised, controlled, multicentre, open-label, parallel-group study (NUTRIREA-2 trial) done at 44 French intensive-care units (ICUs), adults (18 years or older) receiving invasive mechanical ventilation and vasopressor support for shock were randomly assigned (1:1) to either parenteral nutrition or enteral nutrition, both targeting normocaloric goals (20–25 kcal/kg per day), within 24 h after intubation. Randomisation was stratified by centre using permutation blocks of variable sizes. Given that route of nutrition cannot be masked, blinding of the physicians and nurses was not feasible. Patients receiving parenteral nutrition could be switched to enteral nutrition after at least 72 h in the event of shock resolution (no vasopressor support for 24 consecutive hours and arterial lactate <2 mmol/L). The primary endpoint was mortality on day 28 after randomisation in the intention-to-treat-population. This study is registered with ClinicalTrials.gov, number NCT01802099. After the second interim analysis, the independent Data Safety and Monitoring Board deemed that completing patient enrolment was unlikely to significantly change the results of the trial and recommended stopping patient recruitment. Between March 22, 2013, and June 30, 2015, 2410 patients were enrolled and randomly assigned; 1202 to the enteral group and 1208 to the parenteral group. By day 28, 443 (37%) of 1202 patients in the enteral group and 422 (35%) of 1208 patients in the parenteral group had died (absolute difference estimate 2·0%; [95% CI −1·9 to 5·8]; p=0·33). Cumulative incidence of patients with ICU-acquired infections did not differ between the enteral group (173 [14%]) and the parenteral group (194 [16%]; hazard ratio [HR] 0·89 [95% CI 0·72–1·09]; p=0·25). Compared with the parenteral group, the enteral group had higher cumulative incidences of patients with vomiting (406 [34%] vs 246 [20%]; HR 1·89 [1·62–2·20]; p<0·0001), diarrhoea (432 [36%] vs 393 [33%]; 1·20 [1·05–1·37]; p=0·009), bowel ischaemia (19 [2%] vs five [<1%]; 3·84 [1·43–10·3]; p=0·007), and acute colonic pseudo-obstruction (11 [1%] vs three [<1%]; 3·7 [1·03–13·2; p=0·04). In critically ill adults with shock, early isocaloric enteral nutrition did not reduce mortality or the risk of secondary infections but was associated with a greater risk of digestive complications compared with early isocaloric parenteral nutrition. La Roche-sur-Yon Departmental Hospital and French Ministry of Health.

We consider the classical problem of building an arbitrage-free implied volatility surface from bid-ask quotes. We design a fast numerical procedure, for which we prove the convergence, based on the Sinkhorn algorithm that has been recently used to solve efficiently (martingale) optimal transport problems.

We present a set of improvements to the OlmoEarth family. These improvements allow us to cut compute costs during training (\\(1.7 \\) reduction in GPU hours required to train our Base models) and inference (\\(2.9\\) reductions in MACs on Sentinel-2 tasks), while maintaining the models' overall performance. All training code is available at github.com/allenai/olmoearth_pretrain.

We consider the classical problem of building an arbitrage-free implied volatility surface from bid-ask quotes. We design a fast numerical procedure, for which we prove the convergence, based on the Sinkhorn algorithm that has been recently used to solve efficiently (martingale) optimal transport problems.

We consider the classical problem of building an arbitrage-free implied volatility surface from bid-ask quotes. We design a fast numerical procedure, for which we prove the convergence, based on the Sinkhorn algorithm that has been recently used to solve efficiently (martingale) optimal transport problems.

Earth observation data presents a unique challenge: it is spatial like images, sequential like video or text, and highly multimodal. We present OlmoEarth: a multimodal, spatio-temporal foundation model that employs a novel self-supervised learning formulation, masking strategy, and loss all designed for the Earth observation domain. OlmoEarth achieves state-of-the-art performance compared to 12 other foundation models across a variety of research benchmarks and real-world tasks from external partners. When evaluating embeddings OlmoEarth achieves the best performance on 15 out of 24 tasks, and with full fine-tuning it is the best on 19 of 29 tasks. We deploy OlmoEarth as the backbone of an end-to-end platform for data collection, labeling, training, and inference of Earth observation models. The OlmoEarth Platform puts frontier foundation models and powerful data management tools into the hands of non-profits and NGOs working to solve the world's biggest problems. OlmoEarth source code, training data, and pre-trained weights are available at \\(https://github.com/allenai/olmoearth_pretrainhttps://github.com/allenai/olmoearth_pretrain\\).