Search

Addressing the issue of inter-class similarity is a challenging task in the research of autonomous driving scene classification, which primarily focuses on learning the distinctive features of targets in real-world complex traffic scenarios with high similarity, and constructing the overall correlation between features for scene classification. To this end, a multi-scale adaptive feature selection network for autonomous driving scene classification is proposed. Initially, a dual multi-scale feature extraction module is utilized for preliminary processing to extract inter-class similar features at different scales. Subsequently, a feature differentiation screening module is designed to complete the screening of scene-similar features, enabling the network to focus more on the typical and easily distinguishable features of different scene categories. Then, the feature screening results and multi-scale feature maps are transferred to the feature fusion classification module for scene classification, and the correlation between scene features is captured. Finally, an adaptive learning algorithm dynamically adjusts the training parameters through the output results, accelerating the network's convergence speed and improving accuracy. The proposed method is compared with existing network methods on three datasets: BDD100k, BDD100k+ and self-made dataset. Compared with the Top2 networks, it leads in accuracy by 3.29%, 5.59% and 12.65% (relatively), respectively. Experimental results demonstrate the effectiveness of the proposed method and its strong generalization capability. The scene classification method presented in this paper aims to learn the typical and easily distinguishable features and their correlations under different complex scene categories, reducing the impact of inter-class similarity among multiple targets, thereby making the scene classification results in real-world traffic scenario datasets more accurate.

Abstract Background Recently, the oral oncobacterium Fusobacterium nucleatum (F. nucleatum), has been linked with ulcerative colitis (UC). Here, we aim to investigate whether Fecal Microbiota Transplantation (FMT) can alleviate UC by restoring gut microbiota and eliminating oral-derived F. nucleatum and virulence factor fadA. Method C57BL/6J mice were randomly divided into a healthy control group (HC), Dextran Sulfate Sodium group (DSS), oral inoculation group (OR), upper FMT group (UFMT), and lower FMT group (LFMT). Disease activity index, body weight, survival rate, and histopathological scores were used to measure the severity of colitis. The function of the intestinal mucosal barrier was evaluated by performing immunohistochemical staining of the tight junction protein Occludin. Real-time PCR was used to assess the relative abundance of the nusG gene and the virulence gene fadA. Cytokine levels were detected by ELISA. Full-length sequencing of 16S rRNA was used to analyze the changes and composition of gut microbiota. Findings Oral incubation of F. nucleatum further exacerbated the severity of colitis and gut dysbiosis. Peptostreptococcaceae, Enterococcaceae, and Escherichia coli were significantly enriched in OR mice. However, LFMT mice showed an obvious decrease in disease activity and were more effective in restoring gut microbiota and eliminating F. nucleatum than UFMT mice. Bacteroidota, Lachnospiraceae, and Prevotellaceae were mainly enriched bacteria in LFMT mice. In addition, Genera such as Lactobacillus, Allobaculum, and Bacteroidales were found negative correlation with TNF-α, IL-1β, and IL-6. Genera like Romboutsia, Escherichia Shigella, Enterococcus, and Clostridium were found positively correlated with TNF-α, IL-1β, and IL-6. Conclusions Oral incubation of F. nucleatum further exacerbates the severity and dysbiosis in DSS-induced colitis mice. Besides, lower tract FMT can ameliorate colitis by restoring the gut microbiota diversity and eliminating F. nucleatum and virulence factor fadA.

Abstract The potential for off-target mutations is a critical concern for the therapeutic application of CRISPR-Cas9 gene editing. Current detection methodologies, such as GUIDE-seq, exhibit limitations in oligonucleotide integration efficiency and sensitivity, which could hinder their utility in clinical settings. To address these issues, we introduce OliTag-seq, an in-cellulo assay specifically engineered to enhance the detection of off-target events. OliTag-seq employs a stable oligonucleotide for precise break tagging and an innovative triple-priming amplification strategy, significantly improving the scope and accuracy of off-target site identification. This method surpasses traditional assays by providing comprehensive coverage across various sgRNAs and genomic targets. Our research particularly highlights the superior sensitivity of induced pluripotent stem cells (iPSCs) in detecting off-target mutations, advocating for using patient-derived iPSCs for refined off-target analysis in therapeutic gene editing. Furthermore, we provide evidence that prolonged Cas9 expression and transient HDAC inhibitor treatments enhance the assay’s ability to uncover off-target events. OliTag-seq merges the high sensitivity typical of in vitro assays with the practical application of cellular contexts. This approach significantly improves the safety and efficacy profiles of CRISPR-Cas9 interventions in research and clinical environments, positioning it as an essential tool for the precise assessment and refinement of genome editing applications.

Acoustic metasurfaces are widely used for noise attenuation due to their outstanding acoustic performance and subwavelength characteristics. The paper introduces a topology-optimized inverse design approach for broadband sound-absorbing metasurfaces, aiming to achieve efficient sound absorption performance across a wide frequency spectrum. By integrating genetic algorithms with the finite element method and driven by objectives such as the absorption frequency range and absorption coefficient, we can transcend the limitations of manual design and fully exploit the design potential of structures. Furthermore, the study investigates the sound absorption mechanism and thermal-viscous dissipation through finite element simulations and experimental validations. The research findings reveal that the proposed broadband sound-absorbing metasurfaces demonstrate excellent sound absorption capabilities across the designated frequency range, boasting an average sound absorption coefficient exceeding 85%. These findings offer fresh perspectives and methodologies for advancing the research and practical application of broadband sound-absorbing metasurfaces. [ABSTRACT FROM AUTHOR]

Background Accessory breast plasma cell mastitis (PCM) is uncommon in clinical practice with rare reports. Objective To investigate the clinical characteristics, diagnostic methods and the traditional Chinese medicine (TCM) clinical efficacy of accessory breast PCM. Methods The clinical diagnosis and treatment data of 15 patients with accessory breast PCM admitted to Beijing Hospital of Traditional Chinese Medicine, Capital Medical University from January 2017 to December 2021 were analyzed retrospectively, the general information, diagnostic methods, treatment methods and outcomes of 15 cases with accessory breast PCM were summarized. Results A total of 15 female patients of childbearing age with accessory breast PCM were included in this paper, with onset within 2-5 years after pregnancy. The etiology was complex and the course of accessory breast PCM was protracted, which was similar to breast PCM in pathogenetic characteristics, using pathological examination results as gold standard for diagnosis. After the combination of internal and external treatment of TCM, the symptoms of 15 patients were relieved, local lesions disappeared, and there was no recurrence after six months of follow-up. Conclusion Accessory breast PCM is rare in clinical practice, which is easy to be misdiagnosed or missed, and difficult to be treated clinically. The combination of internal and external treatment of TCM has a significant effect on accessory breast PCM.

Abstract Durability is one of the technical bottlenecks restricting fuel cell electric vehicle development. As a result, significant time and resources have been invested in research related to this area worldwide. Current durability research mainly focuses on the single cell and stack levels, which is quite different from the usage scenarios of actual vehicles. There is almost no research on developing durability test cycles on the fuel cell system level. This paper proposes a universal model for developing a durability test cycle for fuel cell system based on the China automotive test cycle. Large-scale comparison tests of the fuel cell systems are conducted. After 1000 h test, the output performance degradation of three mass-produced fuel cell system is 14.49%, 9.59%, and 4.21%, respectively. The test results show that the durability test cycle proposed in this paper can effectively accelerate the durability test of the fuel cell system and evaluate the durability performance of the fuel cell system. Moreover, the methodology proposed in this paper could be used in any other test cycles such as NEDC (New European Driving Cycle), WLTC (Worldwide Harmonized Light Vehicles Test Procedure), etc. And it has comprehensive application value and are significant for reducing the cost of durability testing of fuel cell systems and promoting the industrialization of fuel cell electric vehicles.

Online estimation of the state of health (SOH) of lithium-ion batteries (LIB) is crucial for the security and stability operation of battery management systems. In order to overcome the problem such as long training time, large amount of computation, and complex debugging process of the LIB SOH estimation methods based on traditional data-driven, an LIB SOH estimation method based on fusion health factor (HF) and integrated extreme learning machine is proposed. The interval data with a high correlation with the SOH was found by analyzing the dQ/dV and dT/dV curves of the battery. Multi-dimensional HFs are extracted from the interval data, and the indirect HF are obtained by principal component analysis. The stochastic learning algorithm of extreme learning machine is used to establish the nonlinear mapping relationship between indirect HF and SOH. Considering the unstable output of a single model, an integrated extreme learning machine model is proposed. The unreliable output is eliminated by setting credibility evaluation rules for the estimation results, and the estimation accuracy of the model is improved. Finally, the method proposed in this paper is validated using the NASA LIB aging dataset and the LIB aging dataset of Oxford University. The results show that the average absolute percentage error of SOH estimation method proposed is less than 1%, and it has a high accuracy and reliability.