Search

AIM:To observe the different effects of three intervention approaches on related biometric parameters in children diagnosed with latent myopia, and to investigate different control effects on children with latent myopia.METHODS:Prospective cohort concurrent controls trials. A total of 60 cases(120 eyes)of children who were initially diagnosed as latent myopia and untreated previously at ophthalmology clinic of the Affiliated Hospital of Chengde Medical University from December 2021 to May 2022 were recruited. They were randomly divided into three groups, with 20 cases(40 eyes)in group A treated with 0.01% Atropine eye drops, 20 cases(40 eyes)in group B treated with vision training with a flip chart, and 20 cases(40 eyes)in group C treated with esculin and digitalis glycosides eye drops. They were followed-up for 12 mo, and the spherical equivalent(SE), axial length(AL), corneal curvature(CC), accommodative facility(AF), and macular retinal thickness of the three groups of children were compared at 6 and 12 mo post-intervention.RESULTS:Significant statistical differences were found in AL, SE and AF of the three groups of children at 6 and 12 mo(all P0.05). The retinal thickness of the temporal, nasal, inferior and macular fovea of the outer ring at 6 and 12 mo after intervention in the three groups was significantly different from that at the initial diagnosis(all P0.05).CONCLUSION:When it comes to preventing and controlling myopia, 0.01% Atropine is more effective than flip chart training and esculin and digitalis glycosides eye drops. Therefore, the administration of 0.01% atropine and the implementation of flip chart training can effectively slow down the advancement of latent myopia.

The debonding and sliding of the bolt–slurry interface is the main failure form of the full-length bonding anchor system (FLBAS) of earthen sites, so it is urgent to carry out a quantitative study of the force transfer process of the anchorage interface. Based on field test results and existing research results, it was found that the bilinear bond–slip model is in line with the description of the constitutive relationship of the bolt–slurry interface. The whole process of debonding slip is discussed accordingly; the expressions for the slip, the axial strain of the bolt, and the load displacement at the bolt–slurry interface corresponding to the different loading stages are deduced; and the calculations of the ultimate load-carrying capacity and the effective anchorage length are given at the same time. On this basis, the bond–slip model parameters were calibrated by identifying the characteristic points of the bond–slip curve; a multi-parameter cross-comparison validation of the reasonableness of the theoretical analytical model was carried out on the basis of in situ pull-out tests; and the law of the influence of anchor bond length and axial stiffness on the anchorage performance was analyzed. The analytical model proposed in this study is widely applicable to the analysis of force transfer processes at the bolt–slurry interface in the presence of complete debonding phenomena and provides a useful reference for optimizing the design of anchors while minimizing interventions. [ABSTRACT FROM AUTHOR]

Oils are susceptible to oxidation, so encapsulation of them with small molecule gels and proteins through dynamic high-pressure microfluidization is an effective method of protection. In this study, conjugated linoleic acid (CLA)-rich oil-in-water oleogel emulsions (CLA-Og/W) was prepared by dynamic high-pressure microfluidization. By results, when dynamic high-pressure microfluidization pressure was 80 MPa, zeta potential of CLA-Og/W was −38 mV, the droplet size decreased to 91.3 nm. The surface roughness of the CLA-Og/W reduced to a minimum of 2.75 nm at 80 MPa. The surface tension decreased to 38.36 mN/m. The rheological properties showed that the microfluidized CLA-Og/W had favorable viscoelasticity and increased apparent viscosity. In addition, the fatty acid composition of CLA-Og/W did not change significantly compared to CLA-safflower seed oil. After 14 d of storage at 60 °C, the CLA content of CLA-Og/W was 40.19%. Consequently, CLA-Og/W treated with dynamic high-pressure microfluidization had more stable physicochemical properties and better retention of CLA. • Oleogel emulsion was prepared by dynamic high-pressure microfluidization. • Minimum droplet size obtained at 80 MPa processing pressure. • After 14 d of storage at 60 °C, the conjugated linoleic acid was 40.19%. [ABSTRACT FROM AUTHOR]

Black-box adversarial attacks can be categorized into transfer-based and query-based attacks. The former usually has poor transfer performance due to the mismatch between the architectures of models, while the query-based attacks require massive queries and high dimensional optimization variables. In order to solve the above problems, we propose a novel attack framework integrating the advantages of transfer- and query-based attacks, where the framework is divided into two phases: training the adversarial generator and executing the black-box attacks. In the first stage, a generator is trained by the adversarial loss function so that it can output adversarial perturbation, where the latent variables are designed as the input of the generator to reduce the dimension of the optimization variables. In the second stage, based on the trained generator, we further employ a particle swarm optimization algorithm to optimize the latent variables so that the generator can output the perturbation that can achieve a successful attack. Extensive experiments are performed on the ImageNet dataset, and the results demonstrate that the proposed framework can obtain better attack performance compared with a number of the state-of-the-art black-box adversarial attack methods. In addition, we show the flexibility of the proposed framework by extending the experiment for few-pixel attacks.

Background/aims:  Incidence of colorectal cancer is rapidly increasing worldwide. Extracellular superoxide dismutase (EcSOD; SOD3) is an antioxidant enzyme. However, SOD3 roles in colorectal cancer progression remain uncertain., Materials and Methods:  Superoxide dismutase 3 expression was evaluated, and we analyzed clinical relevance of SOD3 expression in colorectal cancer. Subsequently, SOD3 roles in colorectal cancer progression were detected by gain of function experiments. Changes in subcutaneous tumor and liver nodule size after SOD3 overexpression were examined in nude mice. The expression of proliferation marker Ki67 was assessed by immunohistochemical staining., Results:  Supperoxide dismutase 3 was downregulated in colorectal cancer (P <.01). Downregulation of SOD3 was correlated with unfavorable outcomes (P < .05). Superoxide dismutase 3 upregulation limited the proliferative (P <.05), migrative (P <.01) and invasive actions of colorectal cancer cells (P <.01) by suppressing epithelial-mesenchymal transition. Moreover, SOD3 overexpression reduced Ki67 expression (P <.01) and blocked tumor growth (P <01) and liver metastasis (P <.001) in mouse tumor model., Conclusion:  Superoxide dismutase 3 upregulation attenuates tumor growth and liver metastasis in colorectal cancer, suggesting that SOD3 has potential diagnostic and prognostic values regarding colorectal cancer treatment.

Deep neural networks (DNNs) are shown to be vulnerable to universal adversarial perturbations (UAP), a single quasi-imperceptible perturbation that deceives the DNNs on most input images. The current UAP methods can be divided into data-dependent and data-independent methods. The former exhibits weak transferability in black-box models due to overly relying on model-specific features. The latter shows inferior attack performance in white-box models as it fails to exploit the model's response information to benign images. To address the above issues, this paper proposes a novel universal adversarial attack to generate UAP with strong transferability by disrupting the model-agnostic features (e.g., edges or simple texture), which are invariant to the models. Specifically, we first devise an objective function to weaken the significant channel-wise features and strengthen the less significant channel-wise features, which are partitioned by the designed strategy. Furthermore, the proposed objective function eliminates the dependency on labeled samples, allowing us to utilize out-of-distribution (OOD) data to train UAP. To enhance the attack performance with limited training samples, we exploit the average gradient of the mini-batch input to update the UAP iteratively, which encourages the UAP to capture the local information inside the mini-batch input. In addition, we introduce the momentum term to accumulate the gradient information at each iterative step for the purpose of perceiving the global information over the training set. Finally, extensive experimental results demonstrate that the proposed methods outperform the existing UAP approaches. Additionally, we exhaustively investigate the transferability of the UAP across models, datasets, and tasks.