Diarrhea, a life-threatening gastrointestinal disorder linked to dehydration, remains a global health challenge. While Thymus schimperi is popularly employed in Ethiopian folk medicine for diarrheal conditions, its scientific efficacy is unproven. To assess the antidiarrheal effect of 80% methanol extract and solvent fractions of the leaf of Thymus schimperi in mice. The coarse leaf powder was extracted through soaking by utilizing 80% methanol and separated with distilled water and chloroform. The in vivo antidiarrheal effect was investigated via enteropooling, antimotility tests, and castor oil-induced diarrhea models, with five groups of mice comprising five individuals per group. Negative controls were given 2% Tween 80 or distilled water, while the positive controls were treated with 3 mg/kg of loperamide. 80% methanol extract and fractions were given at doses of 100, 200, and 400 mg/kg. Diarrhea onset time, total and wet stool number and weight, the percent decrease in wet stool number, volume and weight of intraluminal items, and percentage of gastrointestinal transit inhibition were documented. Data analysis was performed using SPSS version 25, with significance assessed via one-way ANOVA followed by Tukey's post hoc test. A p-value < 0.05 was considered statistically significant. The finding revealed that the extract, along with its chloroform and aqueous fractions, significantly inhibited castor oil-induced diarrhea at 400 mg/kg (p < 0.001) and 200 mg/kg (p < 0.01). Additionally, the extract and fractions minimized the intraluminal fluid retention, with the chloroform fraction showing significant effects at 400 mg/kg (p < 0.001) and 200 mg/kg (p < 0.00), while the 80% methanol extract and aqueous fraction were effective at all doses. Furthermore, the extract and fractions decreased gastrointestinal transit in a dose-related pattern, demonstrating significant activity at 100 mg/kg (p < 0.05), 200 mg/kg (p < 0.01), and 400 mg/kg (p < 0.001) in contrast to the vehicle-treated group. This finding illustrated that the 80% methanol extract of Thymus schimperi carried the strongest antidiarrheal efficacy, followed by the aqueous fraction, supportive of its traditional application for alleviating diarrhea.
Cattle behavior constitutes important phenotypic information reflecting animals' health status, activity level, and welfare condition, and is therefore of considerable significance for automated monitoring and precision management in smart livestock farming. However, under complex barn conditions, cattle behavior recognition is easily affected by factors such as illumination variation, partial occlusion, background interference, and individual differences, thereby reducing recognition stability and generalization capability. To address these challenges, this study proposes a pose-driven method for cattle behavior recognition in complex barn environments. First, a 16-keypoint annotation scheme suitable for describing bovine posture, termed cow16, was constructed. Based on this scheme, OpenPose was employed to extract heatmaps (HMs) and part affinity fields (PAFs), which were then used to build an intermediate HM/PAF posture representation. Subsequently, this representation was taken as the input to a lightweight convolutional neural network for classifying three behavioral categories: stand, walk, and lying. On this basis, class-imbalance correction during training and a multi-random-seed logits ensemble strategy during inference were further introduced. In addition, knowledge distillation was adopted to transfer knowledge from a high-performance teacher model to a lightweight student model. Experimental results demonstrate that training-stage class-imbalance correction and inference-stage multi-random-seed logits ensembling exhibit strong complementarity; when combined, the AB configuration improves the test-set Macro-F1 by 3.83 percentage points. Moreover, the distilled student model still achieves competitive recognition performance while maintaining 1× inference cost, indicating a favorable trade-off between accuracy and efficiency. This study provides a useful reference for deployment-oriented cattle behavior recognition in smart farming scenarios and offers a lightweight technical basis for subsequent practical applications.
This study evaluates the protective potential of Fondaparinux (Fond), a selective antithrombin-mediated Factor Xa inhibitor, in methotrexate-induced hepatotoxicity. The work explores its ability to correct coagulation imbalance, improve endothelial function, and attenuate oxidative and inflammatory cascades (TLR4/NLRP3, NF-κB p65/IL-1β/MCP-1). Animals were allocated into 4 groups. A control group was given distilled water via the intraperitoneal route (i.p.); an MTX group was given a single intraperitoneal injection of MTX (20 mg/kg) on the seventh experimental day; and two groups received prior prophylactic administration of Fondaparinux (at doses of 5 or 10 mg/kg, intraperitoneally) throughout seven consecutive days before as well as for an additional four-day period following MTX administration. MTX significantly elevated hepatic injury markers (AST, ALT, ALP), induced oxidative stress with depleted antioxidants (SOD, GSH), and activated TLR4/NLRP3 signaling, resulting in upregulation of inflammatory mediators (TNF-α, NF-κB p65, IL-18, IL-1β, MCP-1, caspase-1, iNOS, ICAM-1, MPO) and suppression of IL-10 (p < 0.05). Endothelial dysfunction was evidenced by reduced eNOS. MTX also triggered marked coagulation disturbances, including enhanced Factor Xa-dependent thrombin generation, increased tissue factor, fibrin deposition, and elevated PAI-1. Mitochondrial apoptotic signaling was promoted, as indicated by elevated expression of cytochrome c along with induced caspase-3 and caspase-9 activation . Histologically, MTX caused extensive hepatic damage characterized by periportal fibrosis, inflammatory infiltration, bile duct proliferation, hepatocellular necrosis, vacuolation, and vascular congestion. Fondaparinux pretreatment dose-dependently restored hemostatic balance, improved endothelial function, suppressed oxidative and inflammatory responses, attenuated apoptosis, and markedly ameliorated histopathological alterations. Fondaparinux limits methotrexate-associated liver damage through inhibition of Factor Xa-dependent coagulation pathways while providing antioxidant, anti-inflammatory, anti-apoptotic, and hepatoprotective actions.
Neural fields (NFs) map continuous coordinates to signals such as color or density, but fast high-quality reconstruction from sparse observations remains difficult. Classical Neural Tangent Kernel (NTK) regression gives closed-form fits, yet it is fundamentally linear and cannot accumulate reusable task priors. We develop three algorithms that address these gaps. NTK-KIP learns a distilled support set of coordinates (and optional labels) so that a finite NTK can inpaint large missing regions from little observed data, yielding a compact non-linear representation instead of a raw kernel solve. MetaQuill meta-learns a shared initialization for an INR so that new scenes can be adapted by updating only a small task-specific weight offset, which provides true feature learning and a reusable prior. Finally, MetaQuill-KIP fuses both ideas: it seeds the task with a KIP-style non-linear warm start, then refines only that small offset around the meta-learned initialization. MetaQuill-KIP achieves high-PSNR reconstructions and semantically plausible inpainting under very sparse observations, while requiring only lightweight per-instance adaptation, whereas diffusion-style baselines typically depend on large pretrained generative priors and costly per-image tuning. This shows that NTK-driven neural fields can be made both non-linear and meta-learnable, narrowing the gap between analytic kernels and practical few-shot reconstruction.
Machine learning (ML) and artificial intelligence (AI) offer opportunity and risk in mass trauma response, disasters and crisis. This narrative review synthesizes material from our "AI to the Rescue" panel at the inaugural PreAct Mass Trauma conference in June 2025, integrating relevant literature and the authors' expertise. We examine AI approaches beyond large language models (LLMs), including traditional ML and multimodal systems, while grounding the concept of "AI-made disasters" as a necessary third disaster type alongside Human-made and Natural, supported by emerging evidence of AI-caused psychiatric harm. We present the AI Safety Levels for Mental Health (ASL-MH) framework with six levels - from supportive applications, to autonomous packages, to experimental, high-risk systems - positioned as a practical heuristic for graduated risk governance given the nascent regulatory landscape and the demonstrated fragility of voluntary industry safety commitments. Using the Model for Adaptive Response to Complex Cyclical Disasters (MARCCD) framework, we organize AI applications across four phases: Anticipation, Impact, Adaptation, and Growth & Recovery, with attention to core disaster mental health sequelae and the challenge of differentiating normative distress from psychopathology. Recommendations address research/evidence, governance/regulation, training/literacy, and equity/access. Given our presentation involved live demos of AI applications, we have distilled key elements into this review which cannot be directly shown.
To evaluate the effects of commonly consumed staining beverages on the color stability and surface roughness of two additively manufactured permanent resin restorative materials. Sixty disk-shaped specimens (10 mm × 2 mm) were fabricated from two additively manufactured resin materials (Saremco Print Crowntec and VarseoSmile Crown Plus). Specimens were finished with standardized polishing procedures and color values (CIE L*a*b*) and baseline surface roughness (Ra) were recorded. Samples were randomly assigned to three immersion media (tea, coffee, distilled water; n = 10 per subgroup) and stored at 37 °C for 14 days with daily solution renewal. Color measurements were repeated after 24 h and 14 days, and color differences were calculated using the CIEDE2000 formula (ΔE₀₀). Surface roughness was re-measured after 14 days. Data were analyzed using mixed-model ANOVA and correlation tests (α = 0.05). Staining solution significantly affected color change at both evaluation periods (p < 0.001), while material type had no significant influence (p > 0.05). Tea produced the highest discoloration values, followed by coffee, whereas water groups remained below acceptability thresholds. Color change increased significantly over time in tea and coffee groups. Surface roughness showed only minimal changes after immersion, with Ra values ranging from 0.438 to 0.559 μm at baseline and 0.444-0.567 μm after 14 days, with no significant effects of material or staining solution (p > 0.05). No significant correlation was found between baseline surface roughness and final color change (p > 0.05). A significant positive correlation was observed between the change in surface roughness (ΔRa) and color change (ΔE00) (p < 0.05). Discoloration of additively manufactured permanent resin materials was primarily influenced by exposure to staining beverages. Tea and coffee caused clinically noticeable color changes, while surface roughness showed limited changes during short-term aging. However, a significant positive correlation between surface roughness change and color change suggests that surface degradation may contribute to staining susceptibility.
To investigate the therapeutic effects and molecular mechanisms of berberine (BBR) for non-alcoholic fatty liver disease (NAFLD) concomitant with type 2 diabetes mellitus (T2DM). In vivo, 16 db/db mice were randomly assigned to the model group and the BBR group by a random number table method (n=8), with db/m mice serving as the control group. Mice were given BBR [100 mg/(kg·d)] or distilled water via gavage for 4 weeks. In vitro, 5-aminoimidazole-4-carboxyamide ribonucleoside (AICAR) and compound C were introduced as a AMP-activated protein kinase (AMPK) agonist and an inhibitor, respectively. HepG2 cells were induced with palmitic acid (PA) and high glucose, and the treatment cells received BBR (5 µmol/L), AICAR (0.8 mmol/L) or compound C (10 µmol/L) or a combination of BBR and compound C for 24 h additionally. Biochemical assays and pathological staining were performed to assess lipid and glucose metabolism. qPCR and Western blot analysis were used to evaluate the mRNA and protein expressions related to fatty acids (FA) translation [including FA transport proteins (FATP) 2, FATP5, CD36], FA synthesis [including stearoyl-CoA desaturase 1 (SCD1), sterol regulatory element-binding proteins-1c (SREBP-1c), fatty acid synthase (FASN)], and FA β-oxidation [acyl-CoA synthetase long-chain family member 1 (ACSL1), carnitine palmitoyltransferase (CPT)1A, CPT1B, CPT2, short-chain-acyl-CoA dehydrogenase(SCAD), medium-chain-acyl-CoA dehydrogenase (MCAD), long-chain-acyl-CoA dehydrogenase (LCAD), and very-long-chain-acyl-CoA dehydrogenase (VLCAD), as well as AMPK/Sirtuin 1 (SIRT1)/peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α) pathway. In vivo, compared with the model group, the mice in the BBR group showed lower TG, TC, LDL-C, fasting blood glucose levels and improved insulin sensitivity, as well as reduced lipid accumulation in liver tissues (P<0.05 or P<0.01). In the molecules related to fatty acid metabolism, the mice in the BBR group showed decreased protein expression of FASN and increased expressions of ACSL1 and CPT1A (P<0.05). Additionally, the mRNA expressions of fatp5 and CD36 were decreased, and CPT1A, CPT2, SCAD, LCAD, and VLCAD were increased (P<0.05). AMPK/SIRT1/PGC-1α pathway was activated in the liver of BBR-treated mice (P<0.05 or P<0.01). In vitro, BBR reduced lipid accumulation in HepG2 cells and activated the AMPK/SIRT1/PGC-1α pathway, and these effects were blocked by compound C (P<0.05 or P<0.01). Berberine activates AMPK/SIRT1/PGC-1α pathway, thereby improving fatty acid metabolism, and ultimately exerts therapeutic effects on NAFLD accompanied by T2DM.
Bitter melon (Momordica charantia L.) is a climbing species frequently employed in traditional medicine to treat a wide number of diseases. Soil salinity represents a primary abiotic constraint on plant growth, impacting nearly 50% of irrigated farmland and constantly increasing. Previous studies revealed the individual benefits of allantoin and serotonin, in mitigating abiotic stresses. Therefore, in the current study, we investigated their ability, both individually and combination, ameliorating the damaging consequences of salt stress on the development and physiological and biochemical attributes of bitter melon. A factorial experiment arranged within a completely randomized design, replicated three times, was performed. Two variables were examined: (a) three salinity levels (0, 50 and 100 mM NaCl) and (b) foliar application treatments (control (distilled water), allantoin (1000 µmol), serotonin (100 µmol) and allantoin (1000 µmol) +serotonin (100 µmol)). At the end of the experiment, critical parameters related to morphology, biochemistry, and physiology were measured from leaf and root tissues, and fruit secondary metabolites was extracted. Vegetative traits (plant height, number of leaves, leaf area), and fruit-related parameters (fruit mean weight and yield) of bitter melon were suppressed under increasing salinity stress. Exposure to 100 mM NaCl reduced photosynthetic efficiency (SPAD (soil-plant analysis development), Fv/Fm), relative content of water and potassium and phosphorus concentration in bitter melon leaf, while increasing electrolyte leakage, proline, sodium, chloride, hydrogen peroxide, malondialdehyde and antioxidant enzyme activities. However, the combined application of allantoin and serotonin alleviated these effects by increasing antioxidant enzyme activity, proline accumulation, total soluble sugar, total protein, ascorbic acid, relative content of water, potassium and phosphorus, while decreasing malondialdehyde and hydrogen peroxide and sodium and chloride accumulation. In fruit, the content of p-coumaric acid (14.64%), ferulic acid (10.48%), caffeic acid (15.74%), gallic acid (22.04%), and cinnamic acid (37.77%) increased in plants treated with allantoin + serotonin compared to the control. The highest amount of phenolic acids was produced by plants treated with allantoin and serotonin. The application of allantoin and serotonin had a positive effect on the quality characteristics and yield of bitter melon fruit compared to the control. In summary, the findings suggest that treating bitter melon with of allantoin and serotonin, particularly in combination, effectively alleviated the adverse impacts of salinity by inducing critical modifications in bitter melon's physiological and biochemical processes. Further investigation might be useful for translating these results to other salt-stressed crops.
Miang, a traditional fermented tea produced from Camellia sinensis var. assamica, is of notable cultural and socio-economic relevance in Northern Thailand. Traditionally, the non-filamentous fungi-based process (NFP) in western Lanna uses only young tea leaves, resulting in substantial amounts of mature leaves being discarded as agricultural waste. This study aimed to utilize the mature tea leaves by adapting the filamentous fungi growth-based process (FFP) of eastern Lanna using selected tannin-tolerant microorganisms, including Aspergillus niger MLF3, Cyberlindera rhodanensis P3, and Lactiplantibacillus pentosus A14-6. Study on fermentation dynamics and bioactive compound formation based on a 2-step fermentation process: 3-day solid-state fermentation with A. niger MLF3, followed by 7-day submerged fermentation by co-culture of C. rhodaninsis P3, and L. pentosus A14-6 in 500 mL sterile distilled water at 30 °C. Increased activities of polysaccharide-degrading enzymes and organic acids were clearly observed during solid-state fermentation, while the significant changes in polyphenol, antioxidant, and reducing sugar content in cell-free supernatant (CFS) were found after submerged fermentation. The obtained CFS shows inhibitory effects of 90 ± 2.5% and 95 ± 1.8% on α-glucosidase and α-amylase, respectively. Analysis of CFS by E-tongue and E-nose clearly indicated the influence of microbial mixture on the taste and aroma of the fermented products. These results demonstrate not only a high-yielding strategy for the effective biotransformation of mature tea leaves into functional drink products but also significant implications for reducing agricultural waste.
Aim The main aim of this study is to compare the shear bond strength (SBS) of 3D-printed customised lingual buttons and prefabricated lingual buttons. Methodology Forty-eight extracted maxillary premolar teeth were selected. Teeth were allocated to two groups of 24 each; in one group, customised lingual buttons were fabricated using 3D printing. Prefabricated lingual buttons were bonded to one group, while the other group was bonded with the customised lingual buttons. The samples were preserved in distilled water at 37⁰C for 24 hours. The samples were evaluated for SBS using a universal testing machine, and the adhesive remnant index (ARI) was determined under a stereomicroscope at 10× magnification. Statistical analysis included the Mann-Whitney U test for SBS and the chi-square test for ARI scores. Results The mean SBS of prefabricated and customised 3D-printed lingual buttons was estimated to be 2.331 ± 2.549 MPa and 3.694 ± 2.564 MPa, respectively. The Mann-Whitney U test conducted to analyse the SBS showed a significant difference between the two groups at p < 0.05. ARI scores indicated that prefabricated lingual buttons had an ARI score of 3 predominantly, while customised 3D-printed lingual buttons had a predominant score of 1. This was statistically significant. Conclusion Customised 3D-printed lingual buttons showed superior SBS compared to the prefabricated lingual buttons. Bond failure was noted more at the bracket-adhesive interface in the prefabricated lingual button group, while it was a cohesive failure close to the enamel surface in the case of customised lingual buttons.
Diffusion models (DMs) have achieved state-of-the-art generative performance but suffer from high sampling latency due to their sequential denoising nature. Existing solver-based acceleration methods often face significant image quality degradation under a low-latency budget, primarily due to accumulated truncation errors arising from the inability to capture high-curvature trajectory segments. In this paper, we propose the Ensemble Parallel Direction solver (dubbed as EPD-Solver), a novel ODE solver that mitigates these errors by incorporating multiple parallel gradient evaluations in each step. Motivated by the geometric insight that sampling trajectories are largely confined to a low-dimensional manifold, EPD-Solver leverages the Mean Value Theorem for vector-valued functions to approximate the integral solution more accurately. Importantly, since the additional gradient computations are independent, they can be fully parallelized, preserving low-latency sampling nature. We introduce a two-stage optimization framework. Initially, EPD-Solver optimizes a small set of learnable parameters via a distillation-based approach. We further propose a parameter-efficient Reinforcement Learning (RL) fine-tuning scheme that reformulates the solver as a stochastic Dirichlet policy. Unlike traditional methods that fine-tune the massive backbone, our RL approach operates strictly within the low-dimensional solver space, effectively mitigating reward hacking while enhancing performance in complex text-to-image (T2I) generation tasks. In addition, our method is flexible and can serve as a plugin (EPD-Solverplugin) to improve existing ODE samplers. Extensive experiments demonstrate the effectiveness of EPD-Solver. On validation benchmarks, at the same latency level of 5 NFE, the distilled EPD-Solver achieves state-of-the-art FID scores of 4.47 on CIFAR-10, 7.97 on FFHQ, 8.17 on ImageNet, and 8.26 on LSUN Bedroom, surpassing existing learning-based solvers by a significant margin. On T2I benchmarks, our RL-tuned EPD-Solver significantly improves human preference scores on both Stable Diffusion v1.5 and SD3-Medium. Notably, it outperforms the official 28-step baseline of SD3-Medium with only 20 steps, effectively bridging the gap between inference efficiency and high-fidelity generation.
Neonicotinoids are nicotine-based synthetic insecticides used in agriculture to control plant pests. They are neurotoxic substances that attack the nervous system of insects and can cause paralysis or death. These selective insecticides should have a negligible effect on non-target organisms, including spiders, which are one of the most abundant and diverse natural predators that contribute to the control of pests. Current studies show that selective insecticides such as neonicotinoids have negative effects on non-target organisms. They can have both lethal effects resulting in mortality, and sublethal effects involving various aspects of their lives, e.g. breeding, movement, hunting, ability to defend against predators, and predatory activity. We studied the species-specific responses to neonicotinoid treatment with the active ingredient thiacloprid of two top spider predators coexisting in tree crowns in Europe - respectively, spiders of the genus Philodromus (aureoles group, Philodromidae) and species Anyphaena accentuata (Walckenaer) (Anyphaenidae). Spiders were exposed to field-realistic concentrations of the tested substance, while the control group was treated with distilled water. We compared the species-specific responses of functional response and two components of predation rate: feeding and overkilling. Further, we monitored long-term survival and recovery from paralysis compared to control, and the impact of insecticide residues on predation rate 14 days post-exposure. We found that a one-hour tarsal contact with thiacloprid significantly reduced predation rate in both Anyphaena and Philodromus spiders, although the effects were highly species-specific. In Anyphaena, feeding was inhibited by fresh treatment, whereas Philodromus remained unaffected in this regard. Furthermore, the rate of overkilling significantly decreased under the fresh treatment in both species. The treatment induced a reversible paralysis in Philodromus, whereas it caused significant mortality in Anyphaena. Furthermore, 14 days post-treatment, the insecticide had no significant effect on feeding of either species, while the overkilling in Philodromus remained significantly lower than in the control. Overall, the study demonstrates a species-specific response to thiacloprid among top pest predators that share the same ecological niche in orchards.
The aim of this study was to evaluate the mechanical (surface roughness & microhardness) and optical (color stability) properties of two modern one-shade dental composites following exposure to three different whitening toothpastes. A total of 192 disc-shaped specimens (n = 8; per subgroup) were prepared from two one-shade composites: ONE (Charisma One, Kulzer, Germany) and VITTRA APS (Advanced Polymerization System) Unique (FGM, Germany). Specimens were assigned to four brushing media: Opalescence Whitening, Signal White Now Glossy Shine, Colgate Optic White Expert, and distilled water (control). Discs (2 mm × 10 mm) were fabricated using a Teflon mold, stored at 37 °C for 24 h in tab water. Microhardness (Vickers; 200 g/10 s), surface roughness (contact profilometry), and color parameters (VITA Easyshade V; CIEDE2000) were recorded before and after a toothbrushing simulation of 10,000 cycles. Toothpastes were used as a 1:3 (v/v) slurry. Data were analyzed using two-way ANOVA with Tukey's HSD test (p < 0.05). All surface roughness values increased after toothbrushing; the greatest change was observed in the VW group (0.056 ± 0.017), whereas the smallest change occurred in the OCO group (0.006 ± 0.040) (p < 0.05). In all subgroups, microhardness values decreased following brushing, and the most pronounced reduction was found in the OW group (-10.12 ± 3.24). This decrease was statistically significant when compared with the OS, OC, OCO, and VW groups (p < 0.05). With respect to color stability, only the VW group remained below the clinically acceptable threshold (ΔE₀₀ = 1.8). Nevertheless, no statistically significant differences were detected among the experimental groups. Nor, interestingly, did any group other than VW exhibit a change that fell within the clinically acceptable range. The compositional architecture and formulation of dental composites, in conjunction with toothpaste constituents, may lead to divergent outcomes when subjected to various mechanical tests.
Reasoning large language models are increasingly considered for healthcare-related artificial intelligence applications, but their practical value depends not only on diagnostic accuracy, but also on responsiveness and operational reliability. In this study, we benchmarked six model settings on 1,000 questions from the MedQA dataset: DeepSeek-R1, its distilled 8-billion-parameter local variant DeepSeek-R1:8b, ChatGPT o3-mini-high, and their knowledge-base-augmented counterparts. We evaluated performance across three dimensions: diagnostic accuracy, response latency, and first-attempt connection reliability. DeepSeek-R1 achieved the highest accuracy (89.5%, 95% CI: 87.4-91.2) but showed substantially longer response times (median 26.54 s) and higher connection failure rates (4.6%). ChatGPT o3-mini-high responded faster (median 10.05 s) and showed the most favorable tail-latency profile, but its accuracy (78.2%, 95% CI: 75.5-80.7) was lower than that of DeepSeek-R1. The locally deployed DeepSeek-R1:8b demonstrated markedly stronger connection reliability (failure rate 0.2%, 95% CI: 0.0%-0.5%) but substantially reduced accuracy (55.0%, 95% CI: 51.9%-58.5%). Knowledge-base augmentation did not consistently improve performance; for DeepSeek-R1, it significantly reduced accuracy by 4.36% ( p = 0.0002 ), while no significant benefit was observed for the other models. These findings show that reasoning model performance in medical question answering is best understood as a trade-off among accuracy, latency, connection reliability, and deployment mode, and that retrieval augmentation is not universally beneficial. More broadly, this study provides deployment-relevant benchmarking evidence for evaluating reasoning models in healthcare-related settings, while also indicating the need for richer knowledge resources and more realistic task environments before such systems can be meaningfully assessed for real-world clinical use.
Climate change led to water scarcity becoming the greatest threat to the growth and yield of maize. Despite the application of growth promoters, the synergistic effects of ascorbic and salicylic acid application remain unexplored, regarding the biochemical mechanisms in the maize plant under drought stress conditions. This study aimed to investigate the impact of foliar applications of ascorbic and salicylic acid on inducing drought stress tolerance in maize. The treatments were comprised of drought (WW = well-watered at 85% FC, DS = drought stress at 40% FC), foliar application (NH = no-spray, DWS = distilled water spray, AA2 = ascorbic acid (200 mg/L), SA2 = salicylic acid (200 mg/L), and AA + SA = combined application (100 mg/L + 100 mg/L). Results showed that drought stress decreased morphological attributes such as shoot length (16%) and leaf area index (11%), photosynthetic pigments like total chlorophyll (15%), and carotenoid levels (19%). However, the combined foliar application of AA + SA significantly increased the above-mentioned morphological attributes (44% and 37%) and photosynthetic pigments (47% and 43%) among the other treatments. Drought stress increased the overproduction of stress indicators such as malondialdehyde (MDA) by 26% and hydrogen peroxide (H2O2) by 23%. However, AA + SA application reduced the overproduction of these stress indicators by 36% and 32%, through improving the activity of enzymatic antioxidants (superoxide dismutase by 45% and catalase by 38%) and the accumulation of osmolytes (proline by 46% and total soluble proteins by 20%). Overall, foliar treatment with AA2 and SA2 is a long-term approach for improving maize defense mechanisms, resulting in increased plant health and production under drought-stress environments.
Large language models (LLMs) have shown promise in interpreting clinical free-text like provider notes. There is limited evidence on tabular electronic health record (EHR) tasks. Our objective was to evaluate the accuracy of LLMs on structured EHR administrative tasks using direct prompting, chain-of-thought (CoT) reasoning, and tool-enabled code generation. We evaluated nine LLMs randomly sampling from a real-world sampled dataset of 50,000 emergency department (ED) visits. Tasks were tested across 25 combinations of table sizes (5-25 rows and columns). Models were prompted directly or with CoT reasoning to return numerical answers. In the tool setting, models generated Python code, which was executed to retrieve answers. Accuracy was defined as the proportion of model outputs matching validated references. We also assessed JSON format compliance. Across 32,950 model queries, performance varied by model, task type, and prompting strategy. Direct prompting produced uniformly low accuracies. CoT prompting moderately improved performance, particularly for logical filtering, but results degraded significantly as table size increased. The tool-based strategy substantially improved accuracy. Smaller models and distilled reasoning variants had more frequent formatting and execution errors. In conclusion, for structured EHR tabular data extraction, direct and CoT prompting strategies resulted in limited accuracy and poor scalability, particularly as table size increased. Tool-based prompting, where models generated and executed Python code, achieved higher accuracy and valid output formatting. Structured data tasks in clinical workflows may require hybrid approaches that combine LLMs with code execution to ensure accuracy and consistency.
Humans are remarkably data efficient when adapting to previously unseen conditions, like driving a new car. In contrast, modern robotic control systems, like neural network policies trained using reinforcement learning (RL), are highly specialized for single environments. Because of this overfitting, they are known to break down even under small differences like the simulation-to-reality gap and require system identification and retraining for even minimal changes to the system. Here, we present RAPTOR, a method for training a highly adaptive foundation policy for quadrotor control. Our method enables training a single, end-to-end neural network policy to control a wide variety of quadrotors. We tested 10 different real quadrotors, from 32 grams to 2.4 kilograms, that also differed in motor type (brushed versus brushless), frame type (soft versus rigid), propeller type (two, three, or four blades), and flight controller (PX4, Betaflight, Crazyflie, M5StampFly). We found that a tiny, three-layer policy with only 2084 parameters was sufficient for zero-shot adaptation to a wide variety of platforms. The adaptation through in-context learning was made possible by using a recurrence in the hidden layer. The policy was trained through our proposed meta-imitation learning algorithm, where we sampled 1000 quadrotors and trained a teacher policy for each of them using RL. The 1000 teachers were distilled into a single, adaptive student policy. We found that within milliseconds, the resulting foundation policy adapted zero-shot to unseen quadrotors. We tested the capabilities of the foundation policy under numerous conditions (trajectory tracking, indoor/outdoor, wind disturbance, poking, and different propellers).
The genus Fusarium comprises several plant pathogenic species, some of which are etiological agents of vascular wilt in agave plants, inducing systemic necrosis across the roots, caudex (cone), and foliar tissues. Such disease compromises host vigor and biomass productivity, ultimately reducing the organoleptic quality of fermented and distilled derivatives, including mezcal, tequila, and other distillates. Thus, as part of our ongoing bioprospecting efforts in unexplored areas of Mexico, the chemical study of the organic extract from a solid-state fermentation culture of the manglicolous endophyte fungus Talaromyces islandicus M31, isolated in the Punta Sur Ecological Park in Cozumel Island Biosphere Reserve, Mexico, led to the separation of eight anthraquinone derivatives active against the agave pathogens Fusarium incarnatum (CRT-153 and CRT-197), Fusarium proliferatum (CRT-142), and Fusarium oxysporum (CRT-098 and CRT-214). Among the molecules tested, the bisdihydroanthraquinone (-)-luteoskyrin (6) exhibited the strongest growth-inhibitory activity, which was also measured over a two-month span. Through macromolecule leakage assay and scanning electron microscopy, we found that (-)-luteoskyrin (6) disrupted membrane integrity in F. incarnatum. Furthermore, a simple formulation consisting of inactivated T. islandicus M31 mycelium was tested in vivo to control Fusarium infection in agave. The results suggest that (-)-luteoskyrin (6) acts synergistically with other components, resulting in remarkable anti-Fusarium activity and demonstrating its efficacy and advantages for pest management.
The leaf sheaths of the red sorghum (Sorghum bicolor L.) Moench) are a plant waste, which has a stable red color due to the bioactive 3-deoxyanthocyanidins such as apigeninidin (APG) and luteolinidin (LUT). In the present study, a multiple optimization process was conducted on ultrasound-assisted extraction (UAE) combined with natural deep eutectic solvent (NADES) to maximize the extraction of 3-deoxyanthocyanidins from the leaf sheaths of the red sorghum. The preliminary experiments showed that the contents of APG and LUT were found in ranges of 2.25 ± 0.01-31.59 ± 0.09 mg/g and 1.36 ± 0.01-30.44 ± 0.11 mg/g, with choline chloride: acetic acid (CHAC), providing the highest values compared with choline chloride: glycerol (CHGLY), choline chloride: ethylene glycol (CHEGLY), choline chloride: lactic acid (CHLA), distilled water, methanol and ethanol. The optimization of the prominent NADES (CHAC) component based on the central composite design resulted in an increase in APG and LUT, being 39.15 ± 0.58 mg/g for APG and 37.89 ± 0.86 mg/g for LUT. The optimization of the extraction conditions using the Box-Behnken design promoted the extraction of APG and LUT, being 54.97 ± 0.96 mg/g for APG and 48.62 ± 0.44 mg/g for LUT at the optimum conditions of 29.93 min, 89.86% and 27.95 mL for extraction time, amplitude and NADES content, respectively. The antioxidant activity of 3-deoxyanthocyanidin-rich extract was found to be 1117.07 ± 21.92 µmol TE/g, 837.36 ± 16.60 mmol TE/g and 5151.03 ± 39.44 mmol ISE/g, for ABTS, DPPH and FRAP assays, respectively. CHAC extract exhibited the highest in vitro bioaccessibility, antidiabetic activity, antimicrobial activity and thermal properties compared with the aqueous extract.
The essential oil (EO) of Sideritis L. has attracted great interest due to its pharmacological activities. At the same time, there is significant variability within the type, related, among other things, to the origin of the raw material. The aim of this work was to study the EO chemical composition of Sideritis scardica Griseb. from Bulgaria and Türkiye. The plant material (air-dried above-ground parts) was purchased from herbal and medical stores in Lublin, Poland. The crushed raw material was used for distillation of the EO. Distillation was performed in a Clevenger apparatus. The EO content was expressed in ml per 100 g of air-dried herb. Analysis of the qualitative and quantitative composition of the obtained EO was performed using gas chromatography coupled with a mass spectrometer (450-GC + 240-MS). The antimicrobial activity of the S. scardica EO was evaluated using the broth microdilution method in accordance with the guidelines of the European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines. We have demonstrated that the chemical composition and biological activity of sideritis EO depend on the origin of the raw material. Our results indicate that S. scardica EO can be considered a promising antimicrobial agent.