Yangzhou Vocational University

Lipid metabolism is critical for seed germination, directly impacting their nutritional value as a food raw material. Conventional methods for oil analysis are destructive and fail to determine oil distribution. This study evaluated the feasibility of using low-field nuclear magnetic resonance (LF-NMR) coupled with magnetic resonance imaging (MRI) as a non-destructive approach for monitoring oil changes in germinating oilseeds. Four representative oilseed varieties - herbaceous (peanut, soybean) and woody (camellia, almond) - were investigated to analyze oil changes during germination. The accuracy of LF-NMR was validated against Soxhlet extraction and confocal laser scanning microscopy (CLSM). The results revealed that herbaceous seeds exhibited rapid oil mobilization germination, whereas woody seeds showed slower oil consumption. High correlations were observed between LF-NMR method and conventional method/CLSM imaging method (R2 > 0.9). Notably, MRI-imaging oil ratio demonstrated the highest accuracy in quantifying both oil content and distribution. Greenness evaluation results show that LF-NMR is the greenest method for sample preparation and measurement process. These findings confirm LF-NMR as an effective method for non-destructive monitoring of oil content and distribution during seed germination, which holds significant application potential in areas such as food raw material quality assessment and the optimization of oilseed processing pretreatment. © 2026 Society of Chemical Industry.

Remote sensing object detection (RSOD) models play an increasingly important role in modern remote sensing systems. However, during model delivery, sharing, and deployment, RSOD models face increasing risks of unauthorized redistribution, illegal replication, and intellectual property infringement. To mitigate these threats, this paper proposes a white-box watermarking framework for RSOD models that enables reliable copyright verification while preserving the performance of the primary detection task. Specifically, a gradient-based sensitivity analysis of the detection loss is first performed to adaptively identify model parameters that minimally affect detection performance, which are then selected as watermark carriers. Subsequently, a parameter-ranking-based watermark encoding scheme is developed, where watermark bits are embedded by enforcing relative ordering constraints between parameter pairs. To further improve robustness under practical deployment conditions, an attack-simulation-driven training strategy is introduced, in which common perturbations and watermark removal attacks are simulated during the embedding process. In addition, a stealthiness enhancement strategy based on statistical distribution constraints is designed to maintain consistency between the distribution of watermarked parameters and those of the original model, thereby reducing the risk of watermark exposure and localization. Extensive experiments across multiple RSOD datasets and detection architectures demonstrate that the proposed method achieves a high copyright verification success rate with negligible impact on detection accuracy and exhibits strong robustness and stealthiness against a variety of watermark removal attacks.

Recent analyses of anti-endothelial cell antibodies (AECAs) have renewed interest in the immunological pathways underlying vascular inflammation. Although endothelial injury mediated by AECAs constitutes a clearly defined mechanistic contributor to a subset of vasculitic syndromes, accumulating evidence from virology, autoinflammation, complement biology, molecular genetics, and tissue injury suggests that vasculitis is better conceptualized as a final common pathological endpoint rather than a uniformly immune-mediated disease. In this perspective, we argue that the traditional immune-centric definition of vasculitis fails to encompass this mechanistic diversity. We outline multiple upstream mechanisms — including infection-driven, immune-complex–mediated, autoinflammatory, complement-mediated, monogenic, and injury-induced pathways — and propose a mechanism-informed nomenclature that integrates conventional vessel-size–based classification with pathway-based endotypes. This dual-layer approach may enhance diagnostic precision, improve biomarker interpretation, guide targeted therapeutic strategies, and better align vasculitis terminology with contemporary biological understanding.

Against the backdrop of rural revitalization and the advancement of lifelong education, this study explores how higher vocational colleges can participate in the construction of new rural communities, using Yangzhou Vocational University as a case study. The research focuses on the university's multidimensional practices, including skill training, cultural empowerment, technological innovation, and community service, to analyze how vocational education can promote the sustainable development of rural communities while fostering the formation of a lifelong learning system. The findings suggest that higher vocational colleges play a critical role in bridging urban-rural educational gaps, enhancing rural human capital, and stimulating local economic and social vitality, thereby providing practical pathways for realizing a learning-oriented society and promoting rural modernization.

Deep learning-based methods have achieved promising performance in image watermarking tasks due to their powerful capability to fully exploit the rich information present in images, which is crucial for ensuring watermark robustness. Although existing methods have improved robustness against various distortions, directly using deep neural networks for feature extraction and watermark expansion often introduces irrelevant and redundant features, thereby limiting the watermarking model’s imperceptibility and robustness. To address these limitations, in this paper, we introduce a robust image watermarking framework (GCMark) based on Gated Feature Selection and Cover-Guided Expansion, which consists of two key components: (1) the Dual-Stage Gated Modulation Block (DSGMB) that serves as the core backbone of both the encoder and decoder, adaptively suppressing irrelevant or redundant activations to enable more precise watermark embedding and extraction; (2) the Cover-Guided Message Expansion Block (CGMEB), which exploits the cover image’s structural features to guide watermark message expansion. By promoting a structurally consistent and well-balanced fusion between the watermark and host image, this implicit symmetry facilitates stable watermark propagation and enhances resilience against various distortions without introducing noticeable visual artifacts. Extensive experimental results demonstrate that the proposed GCMark framework outperforms existing methods with respect to both robustness and imperceptibility.

BackgroundThis study examines mental health services among college students at risk of suicide within Chinese collectivist culture, focusing on the subgroup heterogeneity of help-seeking willingness as outlined by Andersen’s model. By incorporating social learning theory, it highlights the significant role of informal social networks in help-seeking decisions in this cultural context, thereby expanding Andersen’s model applicability in suicide prevention and providing evidence for targeted campus psychological intervention strategies.MethodsA survey of 2,447 Chinese students applied Andersen’s Model, using latent profile analysis to identify help-seeking patterns shaped by suicide risk, depression, suicide stigma, social support, and mental health service availability and accessibility. ANCOVA and PROCESS were used to analyze group differences in intentions and the role of others’ positive experiences.ResultsFour help-seeking groups were identified: Moderate Risk-Resource Limited (19.2%), High-Risk Isolated (4.9%), Low Risk-Stable (55.4%), and Resource Advantage (20.6%). Help-seeking intentions varied significantly (F = 79.16, p < 0.001, partial η² = 0.089), increasing progressively from the High-Risk Isolated to the Resource Advantage Group (ps < 0.01). The perception of others’ positive help-seeking experiences significantly boosted intentions, with varying effects across groups. The High-Risk Isolated Group had the lowest intentions, while the Resource Advantage Group had the highest. Under low exposure to positive experiences, the Moderate Risk-Resource Limited Group showed significantly higher intentions than the Low Risk-Stable Group; however, no significant difference was found between these groups in high exposure.ConclusionsSuicidal college students show diverse traits in predisposing, enabling, and need factors under Andersen’s model, with help-seeking intentions varying significantly across groups and affected by others’ positive experiences. The findings underscore the necessity for targeted, layered interventions to improve campus mental health services while also highlighting the importance of fostering a supportive environment and encouraging colleges to facilitate the sharing of help-seeking experiences to enhance students’ willingness to seek help.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12888-025-07771-7.