Search

The submarine pipe stopper is prone to failure at large pressure and high temperature conditions. This work focuses on the submarine pipe stopper’s sealing failure caused by excessive shear stress and high temperature. First, the working principle of the sealing device in the stopper was analyzed and its performance was verified through land and sea trials. Then a comprehensive evaluation model fit for the large working pressure was established by analyzing the rubber cylinder’s sealing performance and strength. Next, the model was modified by simulation using the nitrile rubber (NBR) cylinder and the fluororubber (FKM) cylinder were used as examples to revise the comprehensive evaluation model while taking temperature into account. The maximum contact stress and the maximum shear stress of the NBR rubber cylinder at 100°C increased by 28% and 30.62% compared to 25°C, and that of the FKM rubber cylinder increased by 24% and 26.12%. In comparison to 25°C, the maximum contact stress and maximum shear stress of the NBR rubber cylinder increased by 28% and 30.62% at 100°C, while the FKM rubber cylinder had an increase of 24% and 26.12%. At 100°C, the NBR rubber cylinder exhibits a maximum contact stress of 4.71 MPa and a maximum shear stress of 1.18 MPa, whereas the FKM rubber cylinder displays maximum contact and shear stresses of 5.09 and 2.44 MPa. Finally, the rubber cylinder’s maximum working pressure at different temperatures was derived, and the influence of its parameters was discussed. The maximum working pressure at 100°C is only 21% of what the rubber cylinder can bear at 25°C. This work is of great significance for accurately evaluating the rubber cylinder’s sealing performance in the submarine pipe stopper and provides new ideas for the rubber cylinder’s design at high-temperature conditions.

It’s important to improve the sealing performance of the inner packer in deepwater pipelines, which is an emergency sealing and oil control equipment. The deformation characteristics and sealing stress conditions of the packer’s sealant cylinder are analyzed, and the hyperelastic constitutive model is determined based on uniaxial tensile experiments. The effective contact stress and effective contact range between the rubber cylinder and the pipe wall are the main parameters for evaluating the sealing performance. Through simulation, it was found that the rubber cylinder on the force application side plays a major sealing role, and severe stress can lead to shoulder protrusions. The location of fatigue damage on the rubber cylinder’s shoulder is consistent with the sea trial. Therefore, special-shaped rubber cylinders with different structures are designed to address the shoulder protrusion. The simulation shows that the sealing performance of the right-angled trapezoidal double-layer rubber cylinder with force applied on the slope side(S-RARC) is significantly better than the original rubber cylinder(ORC), and it does not produce shoulder protrusions.

Purpose: To evaluate the safety and efficacy of Innospring®stent, a novel self-expanding interwoven nitinol stent, in treating femoropopliteal atherosclerotic lesions.Methods: A prospective, single-center, single-arm, first-in-human study enrolled 15 patients (mean age 73.1 years; 13 men) to evaluate the safety and efficacy of the Innospring®stent monitored by core laboratories. The inclusion criteria were claudication or ischemic rest pain, de novo lesions or nonstented restenosis, >70% stenosis, lesion length <20 cm, and a reference vessel diameter of 4–7 mm. The primary safety endpoint was 30-day major adverse events. The primary efficacy end point was stent patency at 12 months. Follow-up evaluations were conducted at 30 days, 6 months, and 12 months.Results: The lesion length was 6.1 ± 3.5 mm. Fourteen (93.3%) patients had lesions of the superficial femoral artery and 3 (20.0%) patients had lesions of the popliteal artery. Nine (60.0%) patients had moderate-to-severe calcified lesion. Technical and procedural success was 100%. No patients experienced major adverse events in the first 30 days. The Rutherford category showed significant and sustained improvement at 6 and 12 months. The 12-month follow-up radiographs obtained in 13 patients confirmed the absence of stent fractures in 100% of examinations. The cumulative primary stent patency rate at 6 and 12 months were 93.3% and 84.6%, respectively.Conclusion: Stenting of the superficial femoral and popliteal arteries using the Innospring®stent is safe and effective. This competing interwoven nitinol stent may provide superior stent integrity and fracture-resistance as well as serve areas under extreme mechanical stress.Clinical Impact Endovascular recanalization is a widely accepted and recommended treatment for symptomatic peripheral artery diseases. The Innospring® stent is a novel self-expanding interwoven stent containing eight nitinol wires with additional radial force, fracture-resistance, and visibility under fluoroscopy. This first-in-human study using the Innospring® stent in patients with femoropopliteal occlusive disease reported that stenting of the superficial femoral and popliteal arteries using the Innospring® stent is safe and effective. This competing interwoven nitinol stent may provide an impressive stent integrity and fracture-resistance as well as serve areas under extreme mechanical stress.

Effective solid waste disposal necessitates suitable combination and clean conversion technologies to maximize resource utilization. Thus, this study investigates the behavior of co-pyrolysis in municipal solid waste (MSW) and food waste (FW) via thermogravimetric analysis, aiming to discern synergistic effects and optimize waste to energy conversion. The impacts of heating rate and blending ratios were evaluated under an inert N2atmosphere, within a temperature range of 30–900℃, during single and combined pyrolysis processes respectively. Results revealed slight variations in weight loss patterns and thermal decomposition behavior. When heating rate increased from 10 to 30℃.min−1, decomposition process delayed, resulting in higher initial (Ti), peak (TP), and final (Tf) temperatures due to inadequate heat transfer. As food waste increased, mass loss rose and favorable synergistic effect was observed at 20 % FW during medium temperatures, while other blends showed inhibitory effects. The pyrolytic index (D) rose nonlinearly from 6.29 × 10−7to 7.72 × 10−7with a higher FW proportion, peaking at 20 % FW, suggesting an optimal MSW and FW mixing ratio. This blend also exhibited the lowest activation energy, indicating a more efficient thermal decomposition process. Those outcomes offer crucial insights for optimizing co-pyrolysis systems and provide essential information for co-disposal of MSW and FW, potentially reducing pollutant emissions and enhancing waste to energy conversion.

Recovering intermediate missing GPS points in a sparse trajectory, while adhering to the constraints of the road network, could offer deep insights into users’ moving behaviors in intelligent transportation systems. Although recent studies have demonstrated the advantages of achieving map-constrained trajectory recovery via an end-to-end manner, they still face two significant challenges. First, existing methods are mostly sequence-based models. It is extremely hard for them to comprehensively capture the micro-semantics of individual trajectory, including the information of each GPS point and the movement between two GPS points. Second, existing approaches ignore the impact of the macro-semantics, i.e., the road conditions and the people's shared travel preferences reflected by a group of trajectories. To address the above challenges, we propose a Micro-Macro Spatial-Temporal Graph-based Encoder-Decoder (MM-STGED). Specifically, we model each trajectory as a graph to efficiently describe the micro-semantics of trajectory and design a novel message-passing mechanism to learn trajectory representations. Additionally, we extract the macro-semantics of trajectories and further incorporate them into a well-designed graph-based decoder to guide trajectory recovery. Extensive experiments conducted on sparse trajectories with three different sampling intervals that are respectively constructed from two real-world trajectory datasets demonstrate the superiority of our proposed model.

Online convex optimization in distributed systems has shown great promise in collaboratively learning on data streams with massive learners, such as in collaborative coordination in robot and IoT networks. When implemented in communication-constrained networks like robot and IoT networks, two critical yet distinct objectives in distributed online convex optimization (DOCO) are minimizing the overall regret and the communication cost. Achieving both objectives simultaneously is challenging, especially when the number of learners $n$n and learning time $T$T are prohibitively large. To address this challenge, we propose novel algorithms in typical adversarial and stochastic settings. Our algorithms significantly reduce the communication complexity of the algorithms with the state-of-the-art regret by a factor of $\mathcal {O}(n^{2})$O(n2) and $\tilde{\mathcal {O}}(\sqrt{nT})$O˜(nT) in adversarial and stochastic settings, respectively. We are the first to achieve nearly optimal regret and communication complexity simultaneously up to polylogarithmic factors. We validate our algorithms through experiments on real-world datasets in classification tasks. Our algorithms with appropriate parameters can achieve $90\%\sim 99\%$90%∼99% communication saving with close accuracy over existing methods in most cases. The code is available at https://github.com/GGBOND121382/Communication-Efficient_Regret-Optimal_DOCO.

Federated learning (FL) enables distributed participants to collaboratively train a machine learning model without accessing to their local data. In FL systems, the selection of training samples has a significant impact on model performances, e.g., selecting participants whose datasets have low-quality samples, features would result in low accuracy, unstable models. In this work, we aim to solve the problem that selects a collection of high-quality training samples for a given FL task under a monetary budget. We propose a holistic design to efficiently select high-quality samples while preserve the privacy of participants’ local data, the server’s label set. We propose an efficient hierarchical sample selection mechanism to select relevant clients, their samples before training for horizontal federated learning (HFL). It uses the determinantal point process (DPP) to select both the statistical homogenous, content diverse clients, samples. Besides, we propose a private set intersection (PSI) based scheme to filter relevant features for the target VFL task. Finally, during training, an erroneous-aware importance based selection is proposed to dynamically select important clients, samples to accelerate model convergence. We verify the merits of our proposed solution with extensive experiments on a real AIoT system with 50 clients. The experimental results validate that our solution achieves accurate, efficient selection of high-quality data, consequently an FL model with a faster convergence speed, higher accuracy.

Modulation of the surface chemistry of air electrodes makes it possible to significantly improve the electrocatalytic performance of solid oxide cells (SOCs). Here, the surface chemistry of BaGd0.8La0.2Co2O6−δ(BGLC) double perovskite is modulated by treatment in an acidic citric acid solution. The treatment leads to corrosion on the surface of BGLC particles, and the effect is dependent on the acidity of the solution. As the acidity of solution is low, Ba cations are selectively dissolved out of the BGLC surface, while as the acidity increases, the corrosion becomes more homogeneous. The Ba surface deficiency remarkably increases the concentration of surface oxygen vacancies and electrocatalytic activity of BGLC. To avoid the loss of Ba-deficient surface during the conventional high temperature sintering process, a sintering-free fabrication route is utilized to directly assemble the Ba-deficient BGLC powder into an air electrode. A single cell with the surface Ba-deficient BGLC electrode shows a peak power density of 1.04 W cm–2at 750 °C and an electrolysis current density of 1.48 A cm–2at 1.3 V, much greater than 0.64 W cm–2and 1.02 A cm–2of the cell with the pristine BGLC, respectively. This work provides a simple and effective surface chemistry modulation strategy for the development of an efficient air electrode for SOCs.

Androgen deprivation therapy (ADT) is the first line of treatment for metastatic prostate cancer (PCa) that effectively delays the tumor progression. However, it also increases the risk of venous thrombosis event (VTE) in patients, a leading cause of mortality. How a pro-thrombotic cascade is induced by ADT remains poorly understood. Here, we report that protein disulfide isomerase A2 (PDIA2) is upregulated in PCa cells to promote VTE formation and enhance PCa cells resistant to ADT. Using various in vitro and in vivo models, we demonstrated a dual function of PDIA2 that enhances tumor-mediated pro-coagulation activity via tumor-derived extracellular vehicles (EVs). It also stimulates PCa cell proliferation, colony formation, and xenograft growth androgen-independently. Mechanistically, PDIA2 activates the tissue factor (TF) on EVs through its isomerase activity, which subsequently triggers a pro-thrombotic cascade in the blood. Additionally, TF-containing EVs can activate the Src kinase inside PCa cells to enhance the AR signaling ligand independently. Androgen deprivation does not alter PDIA2 expression in PCa cells but enhances PDIA2 translocation to the cell membrane and EVs via suppressing the clathrin-dependent endocytic process. Co-recruitment of AR and FOXA1 to the PDIA2 promoter is required for PDIA2 transcription under androgen-deprived conditions. Importantly, blocking PDIA2 isomerase activity suppresses the pro-coagulation activity of patient plasma, PCa cell, and xenograft samples as well as castrate-resistant PCa xenograft growth. These results demonstrate that PDIA2 promotes VTE and tumor progression via activating TF from tumor-derived EVs. They rationalize pharmacological inhibition of PDIA2 to suppress ADT-induced VTE and castrate-resistant tumor progression.

Cuproptosis is a novel copper-dependent form of programmed cell death, displaying important regulatory functions in many human diseases, including cancer. However, the relationship between the changes in mitochondrial viscosity, a key factor associated with cellular malfunction, and cuproptosis is still unclear. Herein, we prepared a phosphorescent iridium (Ir) complex probe for precisely monitoring the changes of mitochondrial viscosity during cuprotosis viaphosphorescence lifetime imaging. The Ir complex probe possessed microsecond lifetimes (up to 1 μs), which could be easily distinguished from cellular autofluorescence to improve the imaging contrast and sensitivity. Benefiting from the long phosphorescence lifetime, excellent viscosity selectivity, and mitochondrial targeting abilities, the Ir complex probe could monitor the increase in the mitochondrial viscosity during cuproptosis (from 46.8 to 68.9 cP) in a quantitative manner. Moreover, through in situfluorescence imaging, the Ir complex probe successfully monitored the increase in viscosity in zebrafish treated with lipopolysaccharides or elescolomol-Cu2+, which were well-known cuproptosis inducers. We anticipate that this new Ir complex probe will be a useful tool for in-depth understanding of the biological effects of mitochondrial viscosity during cuproptosis.

• Cross rolling can separate the broken I-phase particles. • Cross rolling can enhance mechanical strength and eliminate mechanical anisotropy. • Aggregated I-phase particles cause the cracking during tensile process. The microstructure and mechanical performance of the unidirectionally and cross-directionally rolled Mg-8Li-6Zn-1Y (in wt.%) sheets have been investigated and compared. It reveals that after the unidirectional rolling (UR), the broken I-phase particles are aggregated at the α-Mg/β-Li phase interfaces. However, the cross-rolling (CR) process can not only severely break the bulk I-phase, but also cause the obviously uniform distribution of I-phase particles in the matrix phases. Moreover, the average grain size of the CR samples is 3.61 μm and about 50% that of the UR samples. The maximum texture intensities of α-Mg and β-Li phases in the CR samples are slightly stronger than those in the UR samples. Tensile results demonstrate that the CR process can effectively enhance the tensile properties and remarkably reduce the mechanical anisotropy of the alloy. For the UR samples, the yield strength, ultimate tensile strength, and elongation ratio along the rolling direction (RD) are 164 MPa, 198 MPa, and 16.4%, whereas those along the transverse direction (TD) are 157 MPa, 185 MPa, and 22.0%, respectively. For the CR samples, their mechanical properties are basically the same and the mechanical anisotropy is almost eliminated. The yield strength, ultimate tensile strength, and elongation ratio along the cross-rolling direction 1 (CRD1) and 2 (CRD2) are respectively measured to be 181 MPa and 182 MPa, 220 MPa and 218 MPa, 20.6% and 20.7%. Failure analysis indicates that for the UR samples being tensile tested along the RD and TD, micro-cracks are preferentially initiated in the region of aggregated I-phase particles. For the CR samples being tensile tested along both two cross-rolling directions, the initiation of micro-cracks mainly occurs at the I-phase/matrix phase interfaces and in the interior of matrix phases. [Display omitted] [ABSTRACT FROM AUTHOR]

In long-lived room-temperature phosphorescence (RTP) materials, intermolecular interactions have been increasingly acknowledged to play crucial roles, since more and more works have proved their important role in stabilizing the triplet excitons. Herein, we mainly focus on the intermolecular hydrogen bonds and will present a computational study on their regulation of the RTP lifetime of isomer molecules in crystal form and polymorphs. The hydrogen bond is found to have a corresponding relation with phosphorescence lifetime, i.e., an isomerism crystal with a longer RTP lifetime has a larger intermolecular hydrogen-bond contribution. Further investigation on the radiative and nonradiative transition processes illustrates that hydrogen bonds have a considerable impact on the phosphorescent radiative decay rate kp, while nonradiative processes play a weak role in regulating the RTP lifetime in these crystals. Thus, the crystals with a higher intermolecular hydrogen-bond contribution could obtain a longer lifetime mainly by reducing kp. This research may supply a facile way to achieve a long phosphorescence lifetime through enhancing intermolecular hydrogen-bond contribution.

Background: Vascular access (VA) is known to be critical for the survival of patients on maintained hemodialysis (MHD) treatment. However, the association between VA satisfaction and psychiatric state in MHD patients is still not fully elucidated. Thus, the aim of this study is to evaluate the relationship among VA satisfaction, health-related quality of life (HRQOL) and depression in MHD patients with different VA types.Methods: This cross-sectional study was conducted at two dialysis centers with MHD-dependent patients. The Short Form Vascular Access Questionnaire (VAQ) was administered to estimate the level of MHD patients’ satisfaction with their VA. HRQOL was assessed using the Short Form 36 (SF-36) questionnaire. Depression was assessed using the Zung’s self-rating depression scale (SDS).Results: Of the total 252 patients, AVF was used by 84.13%, AVG was used by 2.78%, and TCC was used by 13.09%. There was no significant difference in satisfaction and SDS scores by access type in patients with AVF, AVG, and TCC. However, HRQOL was worst in patients with TCC, and highest in the AVF group. Further analysis showed that VAQ scores in the domains of overall and dialysis-related complications exhibited a negative correlation with HRQOL. And SF-36 HRQOL scores, including the total score, PCS and MCS, were all negatively correlated with SDS scores (p< 0.05). The results of multivariable analyses found that VAQ scores in the domains of overall score and physical symptom, and total score of HRQOL influenced the depression.Conclusions: In the present study, no significant difference in satisfaction scores by access type was found in patients with AVF, AVG, and TCC. The HRQOL score was higher in patients with AVF than in those with AVG or TCC. And the result suggested a negative association between HRQOL and depression. Vascular access satisfaction and HRQOL might be risk factors for the presence of depression in MHD patients.

In this article, we address the distributed state estimation problem for both continuous-time linear time-invariant (LTI) systems and discrete-time LTI systems under switching networks. The observed system is jointly observable, i.e., each agent can only access a part of the measurement output of the observed system and cannot recover the full state by itself. The full state estimation has to be achieved by network communication of neighboring agents. In contrast to existing works, the salient feature of this work is that the developed approach can deal with jointly connected switching networks and thus is more resilient to unreliable communication. First, we propose a new observability decomposition method for linear systems in modal canonical form. Then, we design distributed observers for both the continuous-time system and the discrete-time system. Based on the common Lyapunov function approach, we show that the switched estimation error system is asymptotically stable and thus, the full state estimation can be achieved under jointly connected switching networks.

Digital transformation improves the operational efficiency and perceptual capabilities of businesses and creates a foundation for fulfilling ESG responsibilities. The authors use the theories of technological innovation and dynamic capabilities and select a sample of 4054 listed companies with commercial integration of ESG rating data to explore the mechanisms through which digital transformation affects ESG performance. Dynamic capabilities are further divided into innovation changeability (IT), coordination integration capability (CI), and learning absorption capability (LA) to investigate the mediating effects and summarize the causal path of “digital transformation - dynamic capabilities - ESG performance.” The research focuses on the impact of digital technology on ESG performance with the aim of helping businesses utilize technology more effectively to build dual competitive advantages in digitization and ESG.

In this study, silver nanoparticles (AgNPs) are self-assembled onto the polyamide (PA) pore array through hydrogen bonding, resulting in and optimizing the PA/Ag 3D pore array substrates. The best surface-enhanced Raman scattering (SERS) substrate is obtained with a pore depth of 500 nm in the PA array, 30 nm AgNPs, at a pH of 5.0, and a 24 h assembly time. The SERS performance of the substrates is assessed using rhodamine 6G (R6G) as a probe molecule. The detection limit of the R6G molecule reaches 10–13M, and the relative standard deviation is under 20%, indicating good enhancement ability and reproducibility. Furthermore, label-free detection of pesticide contaminant diquat with a detection limit of 2.69 × 10–9M is achieved using the optimized 3D substrate, which meets environmental monitoring requirements for drinking water. The findings demonstrate that this 3D SERS substrate has promising potential for use and development in the fields of contaminant detection and chemical sensing.

Semantic communication (SemCom), a pioneering paradigm that places emphasis on conveying the meaning of information, faces challenges in constructing background knowledge to drive precise reasoning of semantic coding models. Fortunately, the recent emergence of Generative Artificial Intelligence (GAI) technology is promising to create high-quality content that can be harnessed to assist knowledge construction in SemCom, enhancing the reasoning ability of semantic coding models. In this letter, we propose a GAI-assisted SemCom framework, named Gen-SC, where sufficient samples for training SemCom transceivers are generated using GAI as per user contextual information. In addition, to guide the GAI model in producing contextually relevant content, a discriminator is incorporated into Gen-SC to measure the disparity between generated samples and actual samples. The simulation results demonstrate that the Gen-SC achieves higher semantic accuracy, especially when the original training samples are insufficient, in contrast to traditional SemCom without knowledge enhancement.

Gastric cancer is a leading cause of cancer-related deaths in China. Affecting more than 40% of the world’s population, Helicobacter pyloriis a major risk factor for gastric cancer. While previous clinical trials indicated that eradication of H. pyloricould reduce gastric cancer risk, this remains to be shown using a population-based approach. We conducted a community-based, cluster-randomized, controlled, superiority intervention trial in Linqu County, China, with individuals who tested positive for H. pyloriusing a 13C-urea breath test randomly assigned to receiving either (1) a 10-day, quadruple anti-H. pyloritreatment (comprising 20 mg of omeprazole, 750 mg of tetracycline, 400 mg of metronidazole and 300 mg of bismuth citrate) or (2) symptom alleviation treatment with a single daily dosage of omeprazole and bismuth citrate. H. pylori-negative individuals did not receive any treatment. We examined the incidence of gastric cancer as the primary outcome. A total of 180,284 eligible participants from 980 villages were enrolled over 11.8 years of follow-up, and a total of 1,035 cases of incident gastric cancer were documented. Individuals receiving anti-H. pyloritherapy showed a modest reduction in gastric cancer incidence in intention-to-treat analyses (hazard ratio 0.86, 95% confidence interval 0.74–0.99), with a stronger effect observed for those having successful H. pylorieradication (hazard ratio 0.81, 95% confidence interval 0.69–0.96) than for those who failed treatment. Moderate adverse effects were reported in 1,345 participants during the 10-day treatment. We observed no severe intolerable adverse events during either treatment or follow-up. The findings suggest the potential for H. pylorimass screening and eradication as a public health policy for gastric cancer prevention. Chinese Clinical Trial Registry identifier: ChiCTR-TRC-10000979.

Urban residents' behaviour is the primary cause of carbon emissions; flexible nudge encourages individuals to voluntarily adopt a certain behaviour to cut that down. This study proposed a definition, dimensions, and a theoretical model of voluntary carbon-reducing behaviour. Results from testing 902 valid questionnaires showed that: 1) private voluntary carbon-reducing behaviour is better implemented than public behaviour; 2) significant differences exist in age, marital status, and educational level; 3) psychological variables indirectly affect voluntary carbon-reducing behaviour through intention; 4) flexible nudging has significant moderating roles. Guiding strategies are suggested to encourage urban residents to voluntarily reduce carbon emissions.

Zinc alloys have emerged as promising materials for bone regeneration due to their moderate biodegradation rates. However, the blast release of Zn2+from Zn alloy substrates affects cell behaviors and the subsequent osseointegration quality, retarding their early service performance. To address this issue, extracellular matrix-like hydroxyapatite (HA) nanorods were prepared on Zn-1Ca (ZN) by a combined hydrothermal treatment (HT). HA nanoclusters nucleate on the presetting ZnO layer and grow into nanorods with prolonged HT. HA nanorods protect the ZN substrate from serious corrosion and the corrosion rate is reduced by dozens of times compared with the bare ZN, resulting in a significantly decreased release of Zn2+ions. The synergistic effect of HA nanorods and appropriate Zn2+endow ZN implants with obviously improved behaviors of osteoblasts and endothelial cells (e.g.adhesion, proliferation and differentiation) in vitroand new bone formation in vivo. Our work opens up a promising avenue for Zn-based alloys to improve bone regeneration in clinics.

Background/Objective: To investigate impacts of early postnatal macronutrient intakes on growth and body composition of preterm infants within the first 6 months. Subjects/Methods: One hundred and thirty-three very preterm (VPT) and/or very low birth weight (VLBW) infants were consecutively included. Enteral and parenteral macronutrient intakes during the first 28 days were recorded and average daily intakes were calculated. Growth was measured at birth, term age, and 6 months of corrected age (CA). Body composition was examined by air displacement plethysmograph at term age and 6 months of CA. Associations of nutrient intakes with growth and body composition over time were analyzed using generalized estimating equation. Results: After adjusting for covariates, higher daily protein, lipid, and energy intake during the first 28 days was associated with higher weight at term age for every 1 g/kg/day increment of protein and lipid intake, and every 10 kcal/kg/day increment of energy intake was associated with 0.50 (95% CI 0.04, 0.96), 0.29 (95% CI 0.07, 0.51), and 0.27 (95% CI 0.10, 0.44) higher weight z-score, respectively. Higher protein intake was associated with lower z-score of fat mass (FM, β= −1.88, 95% CI −3.53, −0.23) and percentage of body fat (PBF, β= −2.18, 95% CI −3.98, −0.39) at 6 months of CA, but higher lipid and carbohydrate intake was associated with higher FM and PBF z-scores at 6 months of CA. Conclusions: Macronutrient intakes during the first month of life have impacts on growth and body composition before 6 months of age. Higher daily protein intake is associated with a better growth and healthier body composition for VPT/VLBW infants.

Road extraction from remote-sensing images is of great significance for vehicle navigation and emergency insurance. However, the road information extracted in the remote-sensing image is discontinuous because the road in the image is often obscured by the shadows of trees or buildings. Moreover, due to the scale difference of roads in remote-sensing images, it remains a computational challenge to extract small-size roads from remote-sensing images. To address those problems, we propose a road extraction method based on adaptive global feature fusion (AGF-Net). First, a dilated convolution strip attention (DCSA) module is designed from the encoder–decoder structure. It consists of the dilated convolution and the strip attention module, which adaptively emphasizes relevant features in vertical and horizontal directions. Then, multiple global feature fusion modules (GFFM) in the skip connection are designed to supplement the decoder with road detail features, and we design a multi-scale strip convolution module (MSCM) to implement the GFFM module to obtain multi-scale road information. We compare AGF-Net to state-of-the-art methods and report their performance using standard evaluation metrics, including Intersection over Union (IoU), F1-score, precision, and recall. Our proposed AGF-Net achieves higher accuracy compared to other existing methods on the Massachusetts Road Dataset, DeepGlobe Road Dataset, CHN6-CUG Road Dataset, and BJRoad Dataset. The IoU obtained on these datasets are 0.679, 0.673, 0.567, and 0.637, respectively.

Neural network (NN)-based network intrusion detection systems (NIDSs) are becoming popular these days due to their notable advantages. This article reviews the current application of explainable artificial intelligence techniques and tools for explaining the behavior of the NIDS.

As an essential part of the next-generation communication system, the satellite-terrestrial integrated network (STIN) can provide Internet access and content delivery services for terrestrial users in remote areas. However, more and more content-related traffic requested by users poses challenges to the transmission capability of STINs. Thus, how to provide high-quality content delivery service has become an urgent problem in STINs. In this paper, we leverage the caching capabilities of low earth orbit (LEO) satellites and devise a cooperative caching scheme to achieve low latency and bandwidth-efficient content delivery service. First, considering users' diverse request preferences in different geographical locations, we propose a region features prediction model based on the ridge regression method to update users' preference information. Second, according to the similarity of predicted region features, terrestrial regions covered by LEO satellites are further divided into multiple cooperative areas. Third, we propose a cooperative caching algorithm based on game theory to achieve distributed caching decisions in every single cooperative area. By implementing this cooperative caching scheme, terrestrial users can acquire high-quality content delivery service through STINs. Extensive evaluations show that the proposed caching scheme can significantly reduce content delivery delay while saving valuable bandwidth resources compared with the existing ones.

Mobile-centric AI applications have high requirements for the resource-efficiency of model inference. Input filtering is a promising approach to eliminate redundancy so as to reduce the cost of inference. Previous efforts have tailored effective solutions for many applications, but left two essential questions unanswered: (1) theoretical filterability of an inference workload to guide the application of input filtering techniques, thereby avoiding the trial-and-error cost for resource-constrained mobile applications; (2) robust discriminability of feature embedding to allow input filtering to be widely effective for diverse inference tasks and input content. To answer them, we first formulate the input filtering problem and theoretically compare the hypothesis complexity of inference models and input filters to understand the optimization potential. Then we propose the first end-to-end learnable input filtering framework that covers most state-of-the-art methods and surpasses them in feature embedding with robust discriminability. We design and implement InFi that supports different input modalities and mobile-centric deployments. Comprehensive evaluations confirm our theoretical results and show that InFi outperforms strong baselines in applicability, accuracy, and efficiency. InFi can achieve 8.5× throughput and save 95% bandwidth, while keeping over 90% accuracy, for a video analytics application on mobile platforms.

Vertical Federated Learning (VFL) enables multiple data owners, each holding a different subset of features about a largely overlapping set of data samples, to collaboratively train a global model. The quality of data owners’ local features affects the performance of the VFL model, which makes feature selection vitally important. However, existing feature selection methods for VFL either assume the availability of prior knowledge on the number of noisy features or prior knowledge on the post-training threshold of useful features to be selected, making them unsuitable for practical applications. To bridge this gap, we propose the Federated Stochastic Dual-Gate based Feature Selection (FedSDG-FS) approach. It consists of a Gaussian stochastic dual-gate to efficiently approximate the probability of a feature being selected. FedSDG-FS further designs a local embedding perturbation approach to achieve differential privacy for local training data. To reduce overhead, we propose a feature importance initialization method based on Gini impurity, which can accomplish its goals with only two parameter transmissions between the server and the clients. The enhanced version, FedSDG-FS++, protects the privacy for both the clients’ training data and the server's labels through Partially Homomorphic Encryption (PHE) without relying on a trusted third-party. Theoretically, we analyze the convergence rate, privacy guarantees and security analysis of our methods. Extensive experiments on both synthetic and real-world datasets show that FedSDG-FS and FedSDG-FS++ significantly outperform existing approaches in terms of achieving more accurate selection of high-quality features as well as improving VFL performance in a privacy-preserving manner.

Through carrying out the high-temperature tensile experiments on an as-extruded Mg–11wt%Y alloy at 350 °C, 400 °C, 450 °C, 500 °C and 550 °C, the mechanical behavior and fracture mechanisms at elevated temperatures are investigated and compared. Tensile results show that with the increase of temperature, the yield strength and ultimate tensile strength of the alloy increase at first and then decrease, while that the elongation ratio decreases firstly and then increases. For the sample being tested at 350 °C, the values of yield strength, ultimate tensile strength and the elongation ratio are 188 MPa, 266 MPa and 11%, respectively. At 400 °C, the yield strength and ultimate tensile strength reach the maximum values of, respectively, 198 MPa and 277 MPa, but the elongation ratio is the lowest and its value is only 8%. When the applied temperature is increased to 550 °C, the values of yield strength and ultimate tensile strength, respectively, decrease to 140 MPa and 192 MPa and the elongation ratio increases to 38%. Failure analysis demonstrates that the fracture surfaces of different samples are mainly composed of plastic dimples and exhibit the typical characteristic of ductile fracture. The observation to the fracture side surfaces indicates that at the temperatures of 350 °C and 400 °C, microcracks mainly initiate in the interior of Mg24Y5particles. When the temperatures are 450 °C, 500 °C and 550 °C, the cracks preferentially initiate at the Mg24Y5/α-Mg interfaces.

The development of creative applications and intelligent gadgets requires a secure and straightforward interface with human users. We propose, design, and implement ThumbUp, a smartwatch-based two-factor real-time identification and authentication system in which smartwatch users can identify and authenticate themselves using some simple hand and finger movements, such as thumb-up. ThumbUp leverages the signal from the Inertial Measurement Unit (IMU) in in Commercial-Off-The-Shelf (COTS) smart devices to discover the unique pattern generated by each user's simple gestures using a carefully constructed deep learning model. Smart homes provide a comfortable, safe, and efficient living environment, epecially help the sick and aged. We propose strategies for convenient and reliable control in smart homes with gesture command recognition. We build an Auto-Encoder-based filter that reconstructs the raw data to improve the representation of gesture features. Moreover, we adopt the random forest method to analyse the contextual command correlation. And we employ the authentication system based on smartwatch for personalized command feedback and ensure that illegals cannot use the device. We implement our system and undertake rigorous studies to determine its usefulness and efficiency over a three-month period with 65 users. It achieves a 97% accuracy for user classification and an EER of 0.014 for authentication task with a single simple gesture. And our method achieves 91% accuracy for command recognition and 96% command accuracy with contextual informations. Additionally, we conduct a study of user acceptability of our system and explain how gesture proficiency influences authentication accuracy.