Wenzhou-Kean University

Trigeminal neuralgia (TN) is a debilitating neuropathic facial pain disorder in which current treatments often provide incomplete or poorly tolerated relief. Microglia-driven neuroinflammation in the trigeminal system, particularly Toll-like receptor 4 (TLR4)/myeloid differentiation primary response 88 (MyD88) signaling, is increasingly recognized as a key driver of neuropathic pain, and network pharmacology suggests that Liquiritin, a major licorice flavonoid with anti-inflammatory actions, may target this pathway. We aimed to determine whether Liquiritin alleviates infraorbital nerve injury (IONI)-induced TN-like neuropathic pain by suppressing microglial M1-like polarization via the TLR4/MyD88 pathway, and to characterize its effects on head-withdrawal thresholds, conditioned place preference, inflammatory and pain mediators and TLR4/MyD88 signaling in vivo and in LPS-stimulated BV2 microglia. Adult female ICR mice (8-10weeks) underwent infraorbital nerve injury (IONI) or sham surgery and were randomly assigned to Sham + vehicle, IONI + vehicle, IONI + Liquiritin (200mg/kg, oral) or IONI + Pregabalin (10mg/kg) groups (n=10 per group) treated once daily for 18days. Mechanical allodynia (head-withdrawal thresholds to von Frey stimulation) and conditioned place preference were assessed, trigeminal tissues were analyzed by Western blotting, immunofluorescence and flow cytometry for microglial markers, inflammatory cytokines, pain mediators and TLR4/MyD88, and TLR4 antagonist/agonist administration, LPS-stimulated BV2 microglia and network pharmacology plus molecular docking were used to interrogate Liquiritin's TLR4/MyD88-dependent actions. In IONI mice, Liquiritin significantly attenuated mechanical allodynia and increased conditioned place preference compared with IONI + vehicle, yielding head-withdrawal threshold and conditioned place preference improvements. Network pharmacology identified 92 Liquiritin-related components, 194 candidate targets, and 41 neuropathic-pain-related overlapping genes enriched in Toll-like receptor signaling, and molecular docking showed favorable binding to IL-1β (-11.2kcal/mol), TNF-α (-8.35kcal/mol) and TLR4 (-8.76kcal/mol). Blocking TLR4/MyD88 signaling with LRU alleviated IONI-induced pain behaviors and reduced trigeminal IL-1β, TNF-α, Iba1, CD32, CGRP and TRPV1 expression, whereas TLR4 agonist TEA partially reversed Liquiritin-induced behavioral and molecular changes, supporting pathway involvement. In trigeminal tissues, Liquiritin decreased TLR4 and MyD88 expression and suppressed microglial M1-like markers together with IL-1β, TNF-α, TRPV1 and CGRP; in LPS-stimulated BV2 microglia, Liquiritin (100μM) and LRU (10μg/mL) reduced the production of iNOS, Iba1, CD32 and IL-1β/TNF-α, and serum alanine aminotransferase and aspartate aminotransferase activities were not significantly altered at 200mg/kg, while serum creatinine increased. Liquiritin alleviated IONI-induced TN-like neuropathic pain in mice, concomitant with reduced microglial M1-like activation, decreased IL-1β/TNF-α and TRPV1/CGRP levels, and down-regulation of TLR4/MyD88 signaling in trigeminal tissues. These in vivo, in vitro and in silico data support Liquiritin as a promising neuroinflammation-modulating candidate targeting TLR4/MyD88 in trigeminal neuropathic pain, warranting further pharmacokinetic, long-term safety and translational studies.

Current methods for editing pre-trained models face significant challenges, primarily high computational costs and limited scalability. Task arithmetic has recently emerged as a promising solution, using simple arithmetic operations—addition and negation—based on task vectors which are the differences between fine-tuned and pre-trained model weights, to efficiently modify model behavior. However, the full potential of task arithmetic remains underexplored, primarily due to limited mechanisms for overcoming optimization stagnation. To address this challenge, we introduce the notion of difference vector, a generalized form of task vectors derived from the historical movements during optimization. Using difference vectors as directed perturbations, we proposed the Difference Vector-based Anisotropic Scaling Iterative algorithm (DV-BASI) to enable a continuous optimization process for task arithmetic methods without relying on any additional modules or components. Notably, by leveraging escapability and directional advantages of difference vectors, the average performance on different tasks of the multi-task model merged by DV-BASI may even outperform models individually fine-tuned. Based on this observation, we extend the application of difference vectors to a feasible fine-tuning method for single-task models. On the practical side, DV-BASI allows expressive searching directions with few learnable parameters and forms a scalable framework. We also integrate DV-BASI with task arithmetic methods and advanced optimization techniques to achieve state-of-the-art performance on both supervised and unsupervised evaluation protocols.

Since 2015, KUSA and WKU have jointly developed a robust student exchange program that has engaged more than 2,300 participants. Distinguished by its scale, participant diversity, and institutional composition, the program exemplifies an innovative model of cross-campus collaboration. Its success lies in the joint coordination between the Center for International Studies at both institutions, ensuring consistency in vision and implementation. Students from each campus pursue the exchange for varied academic and cultural motivations, while staff across continents operate cohesively as one team, managing the full cycle of sending and receiving students. Led by Dr. Haina Zhu, Dean of Students at WKU, and Jessica Goldsmith Barzilay, MSW, Assistant Vice President for Global Initiatives at KUSA, this chapter presents reflective insights from leaders who have shaped and sustained this transnational partnership.

The massive low earth orbit (LEO) satellite-empowered Internet of Things (MLS-IoT) network has many advantages including high capacity, low latency, large network coverage, and high reliability. However, the openness and broadcasting on satellite communication links pose huge challenges to protect the security of data delivery, especially to defend against eavesdropping. Leveraging the density of LEO satellite deployment and developing efficient secure communication schemes with the aid of physical layer security (PLS) technique deserve further exploration. To the best of our knowledge, it is an almost untouched issue to employ satellites to transmit jamming signal in MLS-IoT systems since lots of available researches in this line mainly considered cooperative jamming only on ground networks. For this purpose, we propose in this paper a PLS based secure communication enhancement scheme in MLS-IoT network. A jamming scheduling and resource allocation problem is presented for the secrecy rate maximization by dynamically switching the roles of satellites on sub-channels for transmitting private information or jamming signal. An iterative strategy is devised to jointly optimizing jamming scheduling, device association, channel assignment, and power allocation. Numerical results validate that our presented optimization approach could significantly improve the system secrecy rate and achieve the optimal jamming scheduling and resource allocation decisions.

Diabetes-induced osteoporosis significantly elevates the risk of fracture-related disability and mortality. Developing effective therapeutic strategies for diabetic-related bone defects has become a pressing concern in both clinical and research domains. This study innovatively constructs a near-infrared light-responsive (NIR) intelligent hydrogel system (carboxymethyl chitosan/gelatin/black phosphorus@bFGF, CG/BPb), utilizing carboxymethyl chitosan and gelatin as the matrix while integrating polydopamine (PDA)-functionalized black phosphorus nanosheets (BP@PDA) as a controlled-release carrier for basic fibroblast growth factor (bFGF). The CG/BPb hydrogel demonstrated remarkable mechanical strength (up to 25kPa compressive stress at 55% strain) and antioxidant capacity, scavenging 81.1% of ROS and 83.3% of hydroxyl radicals. Under NIR irradiation (1W/cm², 5min), the hydrogel achieved a stable photothermal temperature of 42 ± 1°C, enabling controlled release of bFGF (60% cumulative release within 20min at pH 6.5) and phosphate ions. In vitro, assessments revealed that the hydrogel enhanced osteoblast viability by 85% in scratch assays and upregulated osteogenic genes (ALP, Runx2, and OCN). Additionally, it also promoted M2 macrophage polarization (increased CD206, decreased iNOS) and suppressed osteoclast activity via NFATc1 and MAPK pathways. In vivo, in a diabetic rat calvarial defect model, the CG/BPb + NIR group showed significant bone regeneration, with increases in bone volume fraction (BV/TV) and bone mineral density (BMD), alongside enhanced vascularization (elevated CD31/CD34/α-SMA expression). This innovative strategy, grounded in material design and synergistic biological functions, not only provides a new solution for the treatment of diabetic bone defects but also promotes technological progress in the field of bone tissue engineering, with substantial academic value and practical applications.

Abstract Small object detection in unmanned aerial vehicle (UAV) aerial imagery faces substantial challenges due to small target scales, complex backgrounds, noise interference, and so on. To enhance multi-scale feature representation and detection efficiency, this paper proposes MSEF-YOLO11s. Specifically, we first design a lightweight partial multi-scale (LPMS) module, which effectively aggregates cross-scale information and enhances multi-scale representations in the backbone for small objects. Secondly, to dynamically adjust feature weights and mitigate feature conflicts in the neck, we devise a multi-scale boundary-semantic alignment (MS-BSA) based on adaptive attention, which can further avoid computational redundancy for sufficient fusion. Finally, a lightweight shared detail detection head (LSDDH) replaces the decoupled head structure with shared convolutional layers, resolving the issue of parameter explosion associated with adding a dedicated small object detection head. Experimental results demonstrate the effectiveness of the proposed model. Specifically, compared to the baseline YOLO11s, MSEF-YOLO11s achieves an improvement of 6.6% in mAP50 on the VisDrone2019 test set, with only 4.4M increase in parameters. Furthermore, mAP50 on the TinyPerson test set increases from 22.8% to 28.1%, confirming the model’s strong generalization capability.