Search

Abstract The increasing interest in filter pruning of convolutional neural networks stems from its inherent ability to effectively compress and accelerate these networks. Currently, filter pruning is mainly divided into two schools: norm-based and relation-based. These methods aim to selectively remove the least important filters according to predefined rules. However, the limitations of these methods lie in the inadequate consideration of filter diversity and the impact of batch normalization (BN) layers on the input of the next layer, which may lead to performance degradation. To address the above limitations of norm-based and similarity-based methods, this study conducts empirical analyses to reveal their drawbacks and subsequently introduces a groundbreaking complex hybrid weighted pruning method. By evaluating the correlations and norms between individual filters, as well as the parameters of the BN layer, our method effectively identifies and prunes the most redundant filters in a robust manner, thereby avoiding significant decreases in network performance. We conducted comprehensive and direct pruning experiments on different depths of ResNet using publicly available image classification datasets, ImageNet and CIFAR-10. The results demonstrate the significant efficacy of our approach. In particular, when applied to the ResNet-50 on the ImageNet dataset, achieves a significant reduction of 53.5% in floating-point operations, with a performance loss of only 0.6%.

Background and Aims The diagnostic standard of coronary artery disease (CAD) is coronary angiography (CAG). Since CAG is an invasive procedure underscores the need for identifying non-invasive, effective, and innovative biomarkers. Our study aimed to retrospectively analyze hematological markers for predicting the severity of CAD. Methods and Results Case data were collected from 195 CAD patients admitted to the hospital for CAG. According to Gensini score, patients were divided into mild, moderate, and severe CAD groups. Blood indexes and predictive efficacy of the triglyceride-glucose (TyG) index were retrospectively analyzed. Among 195 CAD patients, 81 had mild CAD, 60 had moderate CAD, and 54 had severe CAD. Sex, fast blood glucose (FBG), TyG index, and high-sensitivity C-reactive protein (hs-CRP) significantly differed among the three groups. The TyG index demonstrated higher values in patients with moderate (9.07[8.62-9.44]) and severe (8.98[8.46-9.45]) CAD compared to those with mild CAD (8.75[8.49-9.14]). The AUC of the TyG index was 0.615 (95% confidence interval (CI): 0.536-0.694, P =.004), with a cut-off value of 8.997, specificity of 0.704, and sensitivity of 0.535. Logistics analysis showed the risk of moderate and severe CAD with an odds ratio (OR) value of 2.595 (95% CI: 1.199-5.619, adjusted P = .016) following regrouping by the TyG index optimal cut-off value of 8.997. The TyG index combined with FBG and hs-CRP had an elevated AUC value, significantly higher than other combinations ( P = .011 and 0.02, respectively). Conclusions The severity of CAD is positively correlated with an increased TyG index value. A combination of TyG, FBG, and hs-CRP has demonstrated improved diagnostic efficiency, suggesting its potential as a novel indicator for predicting and diagnosing CAD progression.

This research focuses on the experimental investigation of the low-cycle fatigue characteristics of marine structural steel. The study aimed to explore the behavior of marine-grade steel under low-cycle fatigue conditions. The experimental parameters include the fatigue life of the material, crack propagation behavior, and a comprehensive analysis of mechanical properties associated with various loading conditions. Based on the experimental results, a low-cycle fatigue crack propagation rate model for marine structural steel plates was established using CTOD as a characterization parameter. The primary objective of this research is likely to enhance the durability and safety of maritime structures, providing valuable technical insights for the field of naval engineering.

The escalating demand for processed foods has led to the widespread industrial use of glucose isomerase (GI) for high-fructose corn syrup (HFCS) production. This reliance on GIs necessitates continual Co2+ supplementation to sustain high catalytic activity across multiple reaction cycles. In this study, Serratia marcescens GI (SmGI) was immobilized onto surfaces of the metal-organic framework (MOF) material MOF (Co)-525 to generate MOF (Co)-525-GI for use in catalyzing glucose isomerization to generate fructose. Examination of MOF (Co)-525-GI structural features using scanning electron microscopy-energy dispersive spectroscopy, Fourier-transform infrared spectroscopy, and ultraviolet spectroscopy revealed no structural changes after SmGI immobilization and the addition of Co2+. Notably, MOF (Co)-525-GI exhibited optimal catalytic activity at pH 7.5 and 70 °C, with a maximum reaction rate (Vmax) of 37.24 ± 1.91 μM/min and Km value of 46.25 ± 3.03 mM observed. Remarkably, immobilized SmGI exhibited sustained high catalytic activity over multiple cycles without continuous Co2+ infusion, retaining its molecular structure and 96.38% of its initial activity after six reaction cycles. These results underscore the potential of MOF (Co)-525-GI to serve as a safer and more efficient immobilized enzyme technology compared to traditional GI-based food-processing technologies.

Abstract Background Multiple myeloma remains incurable despite treatment advancements over the last 20 years. LCAR-B38M Cells in Treating Relapsed/Refractory Multiple Myeloma was a phase 1, first-in-human, investigator-initiated study in relapsed/refractory multiple myeloma conducted at four sites in China. The study used LCAR-B38M chimeric antigen receptor-T cells expressing two B-cell maturation antigen-targeting single-domain antibodies designed to confer avidity, and a CD3ζ signaling domain with a 4-1BB costimulatory domain to optimize T-cell activation and proliferation. This chimeric antigen receptor construct is identical to ciltacabtagene autoleucel. In the LEGEND-2 study (n = 57, Xi’an site), overall response rate was 88%; median (95% CI) progression-free survival and overall survival were 19.9 (9.6–31.0) and 36.1 (26.4–not evaluable) months, respectively; and median follow-up was 25 months. This case study reports on a patient with relapsed/refractory multiple myeloma (λ light chain type) who was treated with LCAR-B38M chimeric antigen receptor T cells in the LEGEND-2 study (Xi’an site); he had received five prior lines of treatment and had extensive extramedullary lesions. Case presentation The patient, a 56-year-old Asian male, received cyclophosphamide (500 mg daily × 3 days) as lymphodepletion therapy and a total dose of 0.5 × 106 chimeric antigen receptor + T cells/kg split into three infusions (days 1, 24, and 84 from June to August 2016). He experienced grade 2 cytokine release syndrome after the first infusion; all symptoms resolved with treatment. No cytokine release syndrome occurred following the second and third infusions. His λ light chain levels decreased and normalized 20 days after the first infusion, and extramedullary lesions were healed as of January 2018. He has sustained remission for 5 years and received no other multiple myeloma treatments after LCAR-B38M chimeric antigen receptor T cell infusion. As of 30 October 2020, the patient is still progression-free and has maintained minimal residual disease-negative (10–4) complete response status for 52 months. Conclusions This case provides support that treatment with LCAR-B38M chimeric antigen receptor T cells can result in long-term disease remission of 5 or more years without disease progression in a heavily pretreated patient with extensive extramedullary disease and no other treatment options.

Underwater autonomous path planning is a critical component of intelligent underwater vehicle system design, especially for maritime conservation and monitoring missions. Effective path planning for these robots necessitates considering various constraints related to robot kinematics, optimization objectives, and other pertinent factors. Sample-based strategies have successfully tackled this problem, particularly the rapidly exploring random tree star (RRT*) algorithm. However, conventional path-searching algorithms may face challenges in the marine environment due to unique terrain undulations, sparse and unpredictable obstacles, and inconsistent results across multiple planning iterations. To address these issues, we propose a new approach specifically tailored to the distinct features of the marine environment for navigation path planning of underwater vehicles, named bidirectional cached rapidly exploring random tree star (BCRRT*). By incorporating bidirectional path planning and caching algorithms on top of the RRT*, the search process can be expedited, and an efficient path connection can be achieved. When encountering new obstacles, ineffective portions of the cached path can be efficiently modified and severed, thus minimizing the computational workload while enhancing the algorithm’s adaptability. A certain number of simulation experiments were conducted, demonstrating that our proposed method outperformed cutting-edge techniques like the RRT* in several critical metrics such as the density of path nodes, planning time, and dynamic adaptability.

This paper presents a straightforward approach for determining the low-cycle fatigue (LCF) crack propagation rate in stiffened plate structures containing cracks. The method relies on both the crack tip opening displacement (CTOD) and the accumulative plastic strain, offering valuable insights for ship structure design and assessing LCF strength. Meanwhile, the LCF crack growth tests for the EH36 steel were conducted on stiffened plates with single-side cracks and central cracks under different loading conditions. The effects of stress amplitude, stress ratio, and stiffener position on the crack growth behavior were examined. Fitting and verifying analyses of the test data were employed to investigate the relationship between CTOD and the crack growth rate of EH36 steel under LCF conditions. The results showed that the proposed CTOD-based prediction method can accurately characterize the LCF crack growth behavior for stiffened plate of EH36 steel for use in ship structures. [ABSTRACT FROM AUTHOR]

Abstract Over the past decade, the feed‐in‐tariff (FIT) subsidy policy of China has driven rapid growth in the photovoltaic power generation (PPG) industry. China now boasts the largest installed capacity of PPG around the world. However, the policy‐driven expansion of the PPG industry has not brought about a simultaneous improvement in quality. PPG costs remain high and it is difficult to sustain PPG competitiveness. Therefore, it is imperative to gradually withdraw from the implementation of photovoltaic subsidies. Using a game theory approach, this study investigates the impact of subsidy exit policy on China's PPG industry. The results show that stimulating research and development (R&D) can be difficult under current uncertain subsidy withdrawal plan. The government should announce the plan of subsidy reduction with clear adjustment as well as corresponding effective dates. In addition, under the condition of government‐controlled prices, the social welfare of the monopoly market is higher than that in the competitive market. In order to achieve the same technological progress and target output, the total cost (including enterprise cost and government expenditure) in competitive market is higher than that in monopoly market, which is caused by repeated R&D investment and technical cost. Technical cooperation could lower costs and increase profits, and is therefore recommended.

ObjectivesThe prognostic significance of acute lymphoblastic leukemia (ALL) patients with central nervous system leukemia (CNSL) at diagnosis is controversial. We aimed to determine the impact of CNSL at diagnosis on the clinical outcomes of childhood B-cell ALL in the South China Children’s Leukemia Group (SCCLG).MethodsA total of 1,872 childhood patients were recruited for the study between October 2016 and July 2021. The diagnosis of CNSL depends on primary cytological examination of cerebrospinal fluid, clinical manifestations, and imaging manifestations. Patients with CNSL at diagnosis received two additional courses of intrathecal triple injections during induction.ResultsThe frequency of CNLS at the diagnosis of B-cell ALL was 3.6%. Patients with CNSL at diagnosis had a significantly higher mean presenting leukocyte count (P = 0.002) and poorer treatment response (P

The traditional study on fatigue strength for ship structures usually focuses on high cycle fatigue and ignores low cycle fatigue. However, given the recent trend towards large-scale ship development, the stress and deformation experienced by ship structures are becoming increasingly significant, leading to greater attention being paid to low cycle fatigue damage. Therefore, experimental and numerical studies on crack propagation behavior of cracked plates under low cycle fatigue loads were carried out in this paper, in order to explain the fatigue crack propagation mechanism. The effect of the stress ratio and maximum applied load on the crack propagation behavior was investigated by conducting experimental research on the cracked plate of AH32 steel. The experimental results show that an increasing maximum applied load and decreasing stress ratio will shorten the fatigue life of the cracked plate. Meanwhile, based on the finite element method, the distribution of the stress–strain field at the crack tip and the effect of crack closure were evaluated. The influencing factors such as the stress ratio and crack length were considered in numerical studies, which provided a new way to study the low cycle fatigue crack propagation behavior.