Chronic obstructive pulmonary disease (COPD) is recognized as one of the three foremost causes of mortality among populations throughout the world. For patients diagnosed with acute exacerbation of COPD (AECOPD), three core clinical features are commonly observed: bronchial spasm, mucus accumulation-induced obstruction, and insufficient compensatory function of the lungs. In clinical practice, airway management measures must abide by a core principle – prioritizing non-invasive over invasive interventions while placing respiratory muscle protection at the core – a standard developed to reduce airway irritation and alleviate related tissue injury. Current clinical practice guidelines support evidence-based pharmacological treatment for AECOPD, yet traditional single-modal intervention methods only bring about temporary symptom relief. These conventional approaches cannot fully address the disease's underlying pathological mechanisms, a limitation that often leads to gradual respiratory function decline and marked impairment of patients' overall quality of life. This review collates cutting-edge clinical studies focusing on multimodal airway management for AECOPD, aiming to break the cycle of over-reliance on single therapies by implementing standardized patient assessments and stratified interventions. Beyond this primary aim, the review also seeks to provide practical clinical care guidance for frontline clinicians, optimize patients' short- and long-term prognosis, and clarify key directions for future research on personalized treatment strategies.

Show more

Drug-induced liver injury (DILI) in adults aged 18–60 years remains understudied despite its clinical heterogeneity and rising incidence. This study aimed to characterize the epidemiology, risk factors, and drug-specific profiles of DILI in this demographic. Utilizing data from the FDA Adverse Event Reporting System (FAERS) (2007–2024), we analyzed 17,464 DILI cases. Four disproportionality methods—Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN), and Multi-item Gamma Poisson Shrinker (MGPS)—were employed to identify high-risk drugs. Time-to-onset (TTO) and gender-specific risks were assessed. Seventeen drugs exhibited significant DILI signals. Rifampicin (ROR=19.33, 95%CI: 16.41–22.77), amoxicillin/clavulanic acid (ROR=16.39, 95%CI: 14.48–18.56), and paracetamol (ROR=10.3,95%CI: 9.61–11.04) showed the strongest associations. Non-steroidal anti-inflammatory drugs (NSAIDs) had the shortest median time-to-onset (7 days). Gender subgroup analysis revealed sex-biased hepatotoxicity, with females disproportionately affected by immunosuppressants and males by antibiotics. This large-scale real-world analysis identifies NSAIDs, antibiotics, and immunosuppressants as critical hepatotoxic threats in younger adults. The findings advocate for targeted hepatic monitoring and updated drug labeling to reflect class-specific latency patterns. Gender-tailored risk mitigation strategies are warranted to address sex-based disparities in DILI susceptibility.

Show more

Lipid nanoparticles (LNPs) exhibit significant potential as highly efficient carriers for nucleic acid therapeutics in the treatment of pulmonary diseases. Nebulized inhalation delivery, which directly targets the lungs through respiratory tract, represents an ideal pulmonary administration route. However, the clinical translation of nebulized LNP-based delivery still faces several critical challenges. Intense shear forces during nebulization impair the structural integrity and stability of LNPs, leading to nucleic acid leakage. The mucin network within the pulmonary mucus layer forms a physical barrier that restricts LNP diffusion, while non-specific phagocytosis by alveolar macrophages further decreases delivery efficiency. In recent years, various design strategies have been developed to address these limitations. Modifying LNP component ratios, optimizing buffer formulations, and functionalizing LNP components have collectively improved stability, enhanced mucus penetration, reduced macrophage uptake, and increased cellular uptake by epithelial cells. This systematic review analyzes the key challenges of nebulized LNP delivery, summarizes recent breakthrough, and outlines future research directions, thereby providing theoretical insights for developing efficient pulmonary nucleic acid delivery systems.

Show more

Analyzing the underlying mechanisms of the brain is an ongoing challenge with extensive application prospects in the field of brain science. With the advancements achieved in computer science, researchers gained a deeper understanding of the function of brain signals via reinforcement learning and deep learning models on supercomputers. This paper focuses on deep learning methods designed for brain signal analysis, respectively discussing the applications of convolutional neural networks(CNNs), recurrent neural networks(RNNs), self-attention models and their mixtures in electroencephalogram(EEG), functional magnetic resonance imaging(fMRI), magnetoencephalogram(MEG), and functional near-infrared spectroscopy(fNIRS). The CNN models have widespread applications, especially in EEG, fMRI while RNN models have unique advantage at dealing with complex temporal data. Self-attention based large transformer models are the hotspots nowadays, general pretrained transformer models can effectively and efficiently handle most cases of brain signal analyzing tasks with enormous computing resources and achieve excellent results.

Show more