Search

Liver tumor segmentation is a critical part in the diagnosis and treatment of liver cancer. While U-shaped convolutional neural networks (UNets) have made significant strides in medical image segmentation, challenges remain in accurately segmenting tumor boundaries and detecting small tumors, resulting in low segmentation accuracy. To improve the segmentation accuracy of liver tumors, this work proposes space pyramid attention (SPA)-UNet, a novel image segmentation network with an encoder-decoder architecture. SPA-UNet consists of four modules: (1) Spatial pyramid convolution block (SPCB), extracting multi-scale features by fusing three sets of dilated convolutions with different rates. (2) Spatial pyramid pooling block (SPPB), performing downsampling to reduce image size. (3) Upsample module, integrating dense positional and semantic information. (4) Residual attention block (RA-Block), enabling precise tumor localization. The encoder incorporates 5 SPCBs and 4 SPPBs to capture contextual information. The decoder consists of the Upsample module and RA-Block, and finally a segmentation head outputs segmented images of liver and liver tumor. Experiments using the liver tumor segmentation dataset demonstrate that SPA-UNet surpasses the traditional UNet model, achieving a 1.0 and 2.0% improvement in intersection over union indicators for liver and tumors, respectively, along with increased recall rates by 1.2 and 1.8%. These advancements provide a dependable foundation for liver cancer diagnosis and treatment.

Objective·To explore the influencing factors of bacterial infection in decompensated stage of hepatitis C cirrhosis, and establish a risk prediction model of nomogram.Methods·A total of 574 patients with decompensated hepatitis C cirrhosis were retrospectively collected from The Third People′s Hospital of Kunming between January 2020 and December 2021, and divided into non-infected and infected groups according to whether bacterial infection occurred. The general information, complications, and laboratory indicators were collected. The variables were screened by univariate analysis, and least absolute shrinkage and selection operator (LASSO) regression, and the nomogram model were constructed and verified by multivariate Logistic regression analysis of influencing factors. The decision curve and clinical impact curve (CIC) were used to evaluate the clinical application value of the model.Results·Bacterial infections occurred in 28.4% (163/574) of the patients, with a total of 191 sites, mainly including spontaneous bacterial peritonitis (86/191) and pulmonary bacterial infections (79/191). Totally 78 strains of pathogens were isolated and cultured, mainly including Klebsiella pneumoniae (15/78) and Escherichia coli (15/78). Multivariate Logistic regression analysis showed that age ≥60 years [odds ratio (OR)=2.054, 95% confidence interval (CI) 1.104‒3.822, P=0.023], female (OR=1.701, 95%CI 1.112‒2.602, P=0.014), ascites (OR=2.386, 95%CI 1.601‒3.557, P=0.000), history of invasive procedures in the last two weeks (OR=2.605, 95%CI 1.368‒4.960, P=0.004), and hospitalization time≥2 weeks (OR=1.629, 95%CI 1.098‒2.416, P=0.015) were independent risk factors for bacterial infection in decompensated hepatitis C cirrhosis patients, while infusing human serum albumin (OR=0.324, 95%CI 0.194‒0.542, P=0.000) and high level of total cholesterol (OR=0.675, 95%CI 0.549‒0.830, P=0.000) were protective factors. The nomogram model was constructed with the above seven influencing factors. Receiver operator characteristic (ROC) curve analysis showed that the area under the curve (AUC) was 0.736 and the sensitivity was 80.4%; and the specificity was 65.1%. Hosmer-lemeshow test showed that the model had a good degree of fit (χ2=9.030, P=0.340). The bootstrap method was used for internal repeated sampling for 1 000 times, the average absolute error was 0.010, the calibration curve and the ideal curve were basically fitted, and the predicted values were in good agreement with the actual values. The decision curve showed that the nomogram model had certain clinical practicability in the high risk threshold range (0.040‒0.715). CIC showed that the nomogram model can be used to forecast the high-risk population in different levels.Conclusion·The nomogram risk prediction model constructed in this study has good predictability, consistency and clinical practicability, and can provide evidence for clinicians to preliminary judge the risk of bacterial infection in patients with decompensated hepatitis C cirrhosis.

Herein, interpenetrating network (IPN) hydrogels were successfully prepared for the first time through integrating two cross-linking reactions in water in one pot, including amine-anhydride reaction and subsequent photopolymerization. First, the polymer containing anhydride groups was prepared via free radical copolymerization of maleic anhydride and acrylamide. Second, the as-prepared polymer was dissolved in water with polyacrylamine hydrochloride, N-isopropylacrylamide, polyethylene glycol dimethacrylate, 2-acrylamido-2-methylpropanesulfonic acid, and photoinitiator to obtain a homogeneous transparent solution. Third, the mixed solution was put in water bath of 50[degrees]C for 1 h to obtain the first network via the amine-anhydride click reaction. Afterward, the second network was formed by photopolymerization. Finally, the swelling and drug release behavior of the hydrogels were studied with isoniazid as a model drug. The results show the IPN hydrogel possesses pH and temperature sensitivity and their controlled-release behavior can be tuned via adjusting pH and temperature. In view of the advantages of amineanhydride reaction and photopolymerization, the strategy described here is of significance for the preparation of IPN hydrogel. KEYWORDS amine-anhydride reaction, interpenetrating network, photopolymerization, smart hydrogels, 1 | INTRODUCTION Polymeric hydrogels are cross-linked hydrophilic network which exhibit the ability to swell in aqueous environments and can absorb and hold a lot of water. [1] Among them, [...]

In target-oriented multi-agent tasks, agents collaboratively achieve goals defined by specific objects, or targets, in their environment. The key to success is the effective coordination between agents and these targets, especially in dynamic environments where targets may shift. Agents must adeptly adjust to these changes and re-evaluate their target interactions. Inefficient coordination can lead to resource waste, extended task times, and lower overall performance. Addressing this challenge, we introduce the regulatory hierarchical multi-agent coordination (RHMC), a hierarchical reinforcement learning approach. RHMC divides the coordination task into two levels: a high-level policy, assigning targets based on environmental state, and a low-level policy, executing basic actions guided by individual target assignments and observations. Stabilizing RHMC's high-level policy is crucial for effective learning. This stability is achieved by reward regularization, reducing reliance on the dynamic low-level policy. Such regularization ensures the high-level policy remains focused on broad coordination, not overly dependent on specific agent actions. By minimizing low-level policy dependence, RHMC adapts more seamlessly to environmental changes, boosting learning efficiency. Testing demonstrates RHMC's superiority over existing methods in global reward and learning efficiency, highlighting its effectiveness in multi-agent coordination. [ABSTRACT FROM AUTHOR]

Unmanned surface pursuit is a complex and challenging engineering problem, especially when conducted by multiple intelligent surface vehicles (ISVs). To enhance the pursuit performance and facilitate strategic interaction during the target pursuit, this paper proposes a novel game theory-based maneuver planning method for pursuit ISVs. Firstly, a specific two-player zero-sum pursuit–evasion game (ZSPEG)-based target-pursuit model is formed. To ensure the vehicles reach a quick consensus, a target-guided relay-pursuit mechanism and the corresponding pursuit payoffs are designed. Meanwhile, under the fictitious play framework, the behavioral pattern and the strategies of the target could be fictitiously learned. Furthermore, mixed-strategy Nash equilibrium (MNE) is employed to determine the motions for the vehicles, the value of which is the best response in the proposed ZSPEG model. Finally, simulations verify the effectiveness of the above methods in multi-ISV surface pursuit. [ABSTRACT FROM AUTHOR]

Background and Objectives: The escalating prevalence of gestational diabetes mellitus (GDM) and the limitations associated with utilizing body mass index (BMI) as a predictive measure underscore the imperative need for identifying an optimal early pregnancy predictor. Such a predictor not only mitigates the risk of GDM but also allows for timely implementation of interventions. Materials and Methods: This meta-analysis aimed to explore the association between visceral adipose tissue (VAT) depth and the risk of GDM. A thorough search of PubMed, Embase, and Web of Science databases was conducted up to 30 September 2023. The analysis employed a random-effects model to assess the relationship between VAT depth and the likelihood of GDM. Results: The inclusion criteria encompassed seven studies involving 1315 women, including 225 diagnosed with GDM. Significantly lower VAT depth was observed in the non-GDM group in comparison to the GDM group (Standardized Mean Difference [SMD]: 0.84; 95% Confidence Interval [CI]: 0.52–1.15; p < 0.001). Substantial statistical heterogeneity was noted among studies (I2 = 72.88%, p = 0.001). Through meticulous sensitivity and subgroup analyses, the source of heterogeneity was identified and thoroughly discussed. Subgroup analyses suggest that different GDM diagnostic criteria and VAT definitions all indicate higher VAT depth in GDM patients during early pregnancy. Conclusions: Our findings propose that, during the first trimester, GDM patients exhibit higher VAT depth compared to non-GDM women, highlighting VAT depth as a potential predictive factor for GDM in early pregnancy. This study contributes valuable evidence to the growing body of knowledge surrounding novel predictors for GDM, emphasizing the importance of early intervention strategies. [ABSTRACT FROM AUTHOR]