Machine learning (ML) models have been developed to identify randomised controlled trials (RCTs) to accelerate systematic reviews (SRs). However, their use has been limited due to concerns about their performance and practical benefits. We developed a high-recall ensemble learning model using Cochrane RCT data to enhance the identification of RCTs for rapid title and abstract screening in SRs and evaluated the model externally with our annotated RCT datasets. Additionally, we assessed the practical impact in terms of labour time savings and recall improvement under two scenarios: ML-assisted double screening (where ML and one reviewer screened all citations in parallel) and ML-assisted stepwise screening (where ML flagged all potential RCTs, and at least two reviewers subsequently filtered the flagged citations). Our model achieved twice the precision compared to the existing SVM model while maintaining a recall of 0.99 in both internal and external tests. In a practical evaluation with ML-assisted double screening, our model led to significant labour time savings (average 45.4%) and improved recall (average 0.998 compared to 0.919 for a single reviewer). In ML-assisted stepwise screening, the model performed similarly to standard manual screening but with average labour time savings of 74.4%. In conclusion, compared with existing methods, the proposed model can reduce workload while maintaining comparable recall when identifying RCTs during the title and abstract screening stages, thereby accelerating SRs. We propose practical recommendations to effectively apply ML-assisted manual screening when conducting SRs, depending on reviewer availability (ML-assisted double screening) or time constraints (ML-assisted stepwise screening).

This study aims to assess the therapeutic effects of probiotic oral therapy in paediatric patients with anorexia nervosa (AN) and to investigate its impact on intestinal flora composition, brain–gut peptide levels and overall clinical outcomes. A retrospective study was conducted involving 100 children diagnosed with AN at Xingtang County People’s Hospital between January 2023 and June 2024. Patients were divided into two groups: a control group (n 50) receiving zinc gluconate oral solution alone and an observation group (n 50) receiving zinc gluconate plus probiotics. Outcome measures included intestinal flora analysis, brain–gut peptide levels (somatostatin (SS) and nitric oxide (NO)), clinical efficacy, serum trace element levels (Ca, Zn and Fe) and prognosis, including recurrence rates 6 months post-treatment. Baseline characteristics were similar between the two groups (P > 0·05). After treatment, the observation group showed significantly higher levels of Bifidobacterium and Lactobacillus and lower levels of Enterobacter compared with the control group (P < 0·05). Additionally, the observation group had lower levels of SS and NO (P < 0·05), indicating improved brain–gut communication. Clinical efficacy was significantly higher in the observation group (P < 0·05), with improved serum trace element levels (P < 0·05 for Ca, Zn and Fe). Furthermore, the recurrence rate 6 months post-treatment was significantly lower in the observation group compared with the control group (P < 0·05). Probiotic supplementation in children with AN effectively modulates intestinal flora, improves brain–gut peptide levels and enhances clinical outcomes.

Aircraft with bio-inspired flapping wings that are operated in low-density atmospheric environments encounter unique challenges associated with the low density. The low density results in the requirement of high operating velocities of aircraft to generate sufficient lift resulting in significant compressibility effects. Here, we perform numerical simulations to investigate the compressibility effects on the lift generation of a bio-inspired wing during hovering flight using an immersed boundary method. The aim of this study is to develop a scaling law to understand how the lift is influenced by the Reynolds and Mach numbers, and the associated flow physics. Our simulations have identified a critical Mach number of approximately $0.6$ defined by the average wing-tip velocity. When the Mach number is lower than 0.6, compressibility does not have significant effects on the lift or flow fields, while when the Mach number is greater than $0.6$, the lift coefficient decreases linearly with increasing Mach number, due to the drastic change in the pressure on the wing surface caused by unsteady shock waves. Moreover, the decay rate is dependent on the Reynolds number and the angle of attack. Based on these observations, we propose a scaling law for the lift of a hovering flapping wing by considering compressible and viscous effects, with the scaled lift showing excellent collapse.

Time-domain characterization of ultrashort pulses is essential for studying interactions between light and matter. Here, we propose and demonstrate an all-optical pulse sampling technique based on reflected four-wave mixing with perturbation on a solid surface. In this method, a weak perturbation pulse perturbs the four-wave mixing signal generated by a strong fundamental pulse. The modulation signal of the four-wave mixing, which is detected in the reflection geometry to ensure a perfect phase-matching condition, directly reflects the temporal profile of the perturbation pulse. We successfully characterized multi-cycle and few-cycle pulses using this method. The reliability of our approach was verified by comparing it to the widely employed frequency-resolved optical gating method. This technique provides a simple and robust method for characterizing ultrashort laser pulses.

A high-energy pulsed vacuum ultraviolet (VUV) solid-state laser at 177 nm with high peak power by the sixth harmonic of a neodymium-doped yttrium aluminum garnet (Nd:YAG) amplifier in a KBe2BO3F2 prism-coupled device was demonstrated. The ultraviolet (UV) pump laser is a 352 ps pulsed, spatial top-hat super-Gaussian beam at 355 nm. A high energy of a 7.12 mJ VUV laser at 177 nm is obtained with a pulse width of 255 ps, indicating a peak power of 28 MW, and the conversion efficiency is 9.42% from 355 to 177 nm. The measured results fitted well with the theoretical prediction. It is the highest pulse energy and highest peak power ever reported in the VUV range for any solid-state lasers. The high-energy, high-peak-power, and high-spatial-uniformity VUV laser is of great interest for ultra-fine machining and particle-size measurements using UV in-line Fraunhofer holography diagnostics.

This study aimed to demonstrate the utilization value of 1PN embryos. The 1PN zygotes collected from December 2021 to September 2022 were included in this study. The embryo development, the pronuclear characteristics, and the genetic constitutions were investigated. The overall blastocyst formation and good-quality blastocyst rates in 1PN zygotes were 22.94 and 16.24%, significantly lower than those of 2PN zygotes (63.25 and 50.23%, respectively, P = 0.000). The pronuclear characteristics were found to be correlated with the developmental potential. When comparing 1PN zygotes that developed into blastocysts to those that arrested, the former exhibited a significantly larger area (749.49 ± 142.77 vs. 634.00 ± 119.05, P = 0.000), a longer diameter of pronuclear (29.81 ± 3.08 vs. 27.30 ± 3.00, P = 0.000), and a greater number of nucleolar precursor body (NPB) (11.56 ± 3.84 vs. 7.19 ± 2.73, P = 0.000). Among the tested embryos, the diploidy euploidy rate was significantly higher in blastocysts in comparison with the arrested embryos (66.67 vs. 11.76%, P = 0.000), which was also significantly higher in IVF-1PN blastocysts than in ICSI-1PN blastocysts (75.44 vs. 25.00%, P = 0.001). However, the pronuclear characteristics were not found to be linked to the chromosomal ploidy once they formed blastocysts.

In summary, while the developmental potential of 1PN zygotes is reduced, our study shows that, in addition to the reported pronuclear area and diameter, the number of NPB is also associated with their developmental potential. The 1PN blastocysts exhibit a high diploidy euploidy rate, are recommend to be clinically used post genetic testing, especially for patients who do not have other 2PN embryos available.

A recent study published in Oryx proposed that the extinct Javan tiger Panthera tigris sondaica may still survive on the Island of Java, Indonesia, based on mitochondrial DNA analysis of a single hair sample collected from a location where a tiger was reportedly encountered. However, upon reanalysing the genetic data presented in that study, we conclude that there is little support for this claim. The sequences of the putative tiger hair and Javan tiger museum specimens generated are not from tiger cytoplasmic mitochondrial DNA but more likely the nuclear pseudogene copies of mitochondrial DNA. In addition, the number of mismatches between the two Javan tiger sequences is unusually high for homologous sequences that are both from tigers, suggesting potential issues with data reliability. The paper provides insufficient details on quality control measures, making it impossible to rule out the possibility that errors were introduced during the analysis. Consequently, it is inappropriate to use the sequences presented in that study to infer the existence of the Javan tiger.

We aimed to evaluate the association of coffee consumption with different additives, including milk and/or sweetener (sugar and/or artificial sweetener), and different coffee types, with new-onset acute kidney injury (AKI), and examine the modifying effects of genetic variation in caffeine metabolism. 194 324 participants without AKI at baseline in the UK Biobank were included. The study outcome was new-onset AKI. During a median follow-up of 11·6 years, 5864 participants developed new-onset AKI. Compared with coffee non-consumers, a significantly lower risk of new-onset AKI was found in coffee consumers adding neither milk nor sugar to coffee (hazard ratio (HR), 0·86; 95 % CI, 0·78, 0·94) and adding only milk to coffee (HR,0·83; 95 % CI, 0·78, 0·89), but not in coffee consumers adding only sweetener (HR,1·14; 95 % CI, 0·99, 1·31) and both milk and sweetener to coffee (HR,0·96; 95 % CI, 0·89, 1·03). Moreover, there was a U-shaped association of coffee consumption with new-onset AKI, with the lowest risk at 2–3 drinks/d, in unsweetened coffee (no additives or milk only to coffee), but no association was found in sweetened coffee (sweetener only or both milk and sweetener to coffee). Genetic variation in caffeine metabolism did not significantly modify the association. A similar U-shaped association was found for instant, ground and decaffeinated coffee consumption in unsweetened coffee consumers, but not in sweetened coffee consumers. In conclusion, moderate consumption (2–3 drinks/d) of unsweetened coffee with or without milk was associated with a lower risk of new-onset AKI, irrespective of coffee type and genetic variation in caffeine metabolism.

In this paper, a brand-new adaptive fault-tolerant non-affine integrated guidance and control method based on reinforcement learning is proposed for a class of skid-to-turn (STT) missile. Firstly, considering the non-affine characteristics of the missile, a new non-affine integrated guidance and control (NAIGC) design model is constructed. For the NAIGC system, an adaptive expansion integral system is introduced to address the issue of challenging control brought on by the non-affine form of the control signal. Subsequently, the hyperbolic tangent function and adaptive boundary estimation are utilised to lessen the jitter due to disturbances in the control system and the deviation caused by actuator failures while taking into account the uncertainty in the NAIGC system. Importantly, actor-critic is introduced into the control framework, where the actor network aims to deal with the multiple uncertainties of the subsystem and generate the control input based on the critic results. Eventually, not only is the stability of the NAIGC closed-loop system demonstrated using Lyapunov theory, but also the validity and superiority of the method are verified by numerical simulations.

Melting and solidification in periodically time-modulated thermal convection are relevant for numerous natural and engineering systems, for example, glacial melting under periodic sun radiation and latent thermal energy storage under periodically pulsating heating. It is highly relevant for the estimation of melt rate and melt efficiency management. However, even the dynamics of a solid–liquid interface shape subjected to a simple sinusoidal heating has not yet been investigated in detail. In this paper, we offer a better understanding of the modulation frequency dependence of the melting and solidification front. We numerically investigate periodic melting and solidification in turbulent convective flow with the solid above and the melted liquid below, and sinusoidal heating at the bottom plate with the mean temperature equal to the melting temperature. We investigate how the periodic heating can prevent the full solidification, and the resulting flow structures and the quasi-equilibrium interface height. We further study the dependence on the heating modulation frequency. As the frequency decreases, we found two distinct regimes, which are ‘partially solid’ and ‘fully solid’. In the fully solid regime, the liquid freezes completely, and the effect of the modulation is limited. In the partially solid regime, the solid partially melts, and a steady or unsteady solid–liquid interface forms depending on the frequency. The interface height can be derived based on the energy balance through the interface. In the partially solid regime, the interface height oscillates periodically, following the frequency of modulation. Here, we propose a perturbation approach that can predict the dependency of the oscillation amplitude on the modulation frequency.

China is still among the 30 high-burden tuberculosis (TB) countries in the world. Few studies have described the spatial epidemiological characteristics of pulmonary TB (PTB) in Jiangsu Province. The registered incidence data of PTB patients in 95 counties of Jiangsu Province from 2011 to 2021 were collected from the Tuberculosis Management Information System. Three-dimensional spatial trends, spatial autocorrelation, and spatial–temporal scan analysis were conducted to explore the spatial clustering pattern of PTB. From 2011 to 2021, a total of 347,495 newly diagnosed PTB cases were registered. The registered incidence rate of PTB decreased from 49.78/100,000 in 2011 to 26.49/100,000 in 2021, exhibiting a steady downward trend (χ2 = 414.22, P < 0.001). The average annual registered incidence rate of PTB was higher in the central and northern regions. Moran’s I indices of the registered incidence of PTB were all >0 (P< 0.05) except in 2016, indicating a positive spatial correlation overall. Local autocorrelation analysis showed that ‘high–high’ clusters were mainly distributed in northern Jiangsu, and ‘low–low’ clusters were mainly concentrated in southern Jiangsu. The results of this study assist in identifying settings and locations of high TB risk and inform policy-making for PTB control and prevention.

In large-scale galaxy surveys, particularly deep ground-based photometric studies, galaxy blending was inevitable. Such blending posed a potential primary systematic uncertainty for upcoming surveys. Current deblenders predominantly depended on analytical modelling of galaxy profiles, facing limitations due to inflexible and imprecise models. We presented a novel approach, using a U-net structured transformer-based network for deblending astronomical images, which we term the CAT-deblender. It was trained using both RGB and the grz-band images, spanning two distinct data formats present in the Dark Energy Camera Legacy Survey (DECaLS) database, including galaxies with diverse morphologies in the training dataset. Our method necessitated only the approximate central coordinates of each target galaxy, sourced from galaxy detection, bypassing assumptions on neighbouring source counts. Post-deblending, our RGB images retained a high signal-to-noise peak, consistently showing superior structural similarity against ground truth. For multi-band images, the ellipticity of central galaxies and median reconstruction error for r-band consistently lie within $\pm$0.025 to $\pm$0.25, revealing minimal pixel residuals. In our comparison of deblending capabilities focused on flux recovery, our model showed a mere 1% error in magnitude recovery for quadruply blended galaxies, significantly outperforming SExtractor’s higher error rate of 4.8%. Furthermore, by cross-matching with the publicly accessible overlapping galaxy catalogs from the DECaLS database, we successfully deblended 433 overlapping galaxies. Moreover, we have demonstrated effective deblending of 63 733 blended galaxy images, randomly chosen from the DECaLS database.