Hasil untuk "Therapeutics. Pharmacology"

Oliver Zahn, Simran Chana

Retrieval-augmented generation stores all content indiscriminately, degrading accuracy as noise accumulates. Parametric approaches compress knowledge into weights, precluding selective updates. Neither mirrors biological memory, which gates encoding based on salience and archives rather than deletes superseded information. We introduce write-time gating that filters incoming knowledge objects using composite salience scores (source reputation, novelty, reliability) while maintaining version chains that preserve prior states. Using real LLM evaluation without oracle access to quality labels, write gating achieves 100 percent accuracy versus 13 percent for ungated stores. The critical finding emerges under distractor scaling: at 8:1 distractor ratios, read-time filtering (Self-RAG) collapses to 0 percent while write gating maintains 100 percent, revealing a structural advantage of write-time over read-time curation. Validation on Wikipedia (20 entities), procedurally generated pharmacology data, and 2026 arXiv papers confirms these findings. The gating advantage scales inversely with parametric memory support: +25pp for Wikipedia, +48pp for post-cutoff arXiv, +65pp for procedural data with zero training knowledge. Signal ablation confirms the method does not depend on oracle-correlated metadata. Write gating matches Self-RAG accuracy at one-ninth the query-time cost.

David C. Norris

Dose-escalation trials in oncology drug development still today typically aim to identify 1-size-fits-all dose recommendations, as arbitrary quantiles of the toxicity thresholds evident in patient samples. In the late 1990s efforts to individualize dosing emerged fleetingly in the oncology trial methods literature, but these have gained little traction due to a nexus of conceptual, technical, commercial, and regulatory barriers. To reduce the activation energy needed for transforming current 1-size-fits-all dose-escalation trial designs to the dose-titration designs required for patient-centered dose individualization, we demonstrate a categorical formulation of dose-escalation protocols that extends readily to allow gradual introduction of dose titration. Central to this formulation is a symmetric monoidal preorder on the accessible states of dose-escalation trials, embodying pharmacologic intuitions regarding dose-monotonicity of drug toxicity and ethical intuitions relating to the therapeutic intent of such trials. A trial protocol that assigns doses to enrolling participants consistently with these intuitions is then a monotone map from this preorder to the sequence of doses being trialed. We illustrate this formulation by reference to the ubiquitous 3+3 dose-escalation design, which despite its many widely discussed flaws remains familiar to oncology trialists and moreover has available an executable specification in Prolog. Remarkably, examined in light of our preorder the 3+3 protocol discloses a new flaw not previously described: a non-monotone dose recommendation. The right Kan extension approximates this protocol from the side of safety, dissolving its triplet cohorts to allow incremental enrollment, and rectifying said non-monotonicity. It also facilitates accelerated enrollment while toxicity assessments remain pending, as well as discretionary dose titration.

Zhongyue Zhang, Jiahua Rao, Jie Zhong et al.

Most human proteins remain undrugged, over 96% of human proteins remain unexploited by approved therapeutics. While structure-based virtual screening promises to expand the druggable proteome, existing methods lack atomic-level precision and fail to predict binding fitness, limiting translational impact. We present AuroBind, a scalable virtual screening framework that fine-tunes a custom atomic-level structural model on million-scale chemogenomic data. AuroBind integrates direct preference optimization, self-distillation from high-confidence complexes, and a teacher-student acceleration strategy to jointly predict ligand-bound structures and binding fitness. The proposed models outperform state-of-the-art models on structural and functional benchmarks while enabling 100,000-fold faster screening across ultra-large compound libraries. In a prospective screen across ten disease-relevant targets, AuroBind achieved experimental hit rates of 7-69%, with top compounds reaching sub-nanomolar to picomolar potency. For the orphan GPCRs GPR151 and GPR160, AuroBind identified both agonists and antagonists with success rates of 16-30%, and functional assays confirmed GPR160 modulation in liver and prostate cancer models. AuroBind offers a generalizable framework for structure-function learning and high-throughput molecular screening, bridging the gap between structure prediction and therapeutic discovery.

Athar Mahmoudi-Nejad, Matthew Guzdial, Pierre Boulanger

Personalized therapy, in which a therapeutic practice is adapted to an individual patient, can lead to improved health outcomes. Typically, this is accomplished by relying on a therapist's training and intuition along with feedback from a patient. However, this requires the therapist to become an expert on any technological components, such as in the case of Virtual Reality Exposure Therapy (VRET). While there exist approaches to automatically adapt therapeutic content to a patient, they generally rely on hand-authored, pre-defined rules, which may not generalize to all individuals. In this paper, we propose an approach to automatically adapt therapeutic content to patients based on physiological measures. We implement our approach in the context of virtual reality arachnophobia exposure therapy, and rely on experience-driven procedural content generation via reinforcement learning (EDPCGRL) to generate virtual spiders to match an individual patient. Through a human subject study, we demonstrate that our system significantly outperforms a more common rules-based method, highlighting its potential for enhancing personalized therapeutic interventions.

Alessandra Scano, Sara Fais, Giuliana Ciappina et al.

Background: <i>Fusobacterium nucleatum</i> is a pathobiont that plays a dual role as both a commensal and a pathogen. The oral cavity typically harbors this anaerobic, Gram-negative bacterium. At the same time, it is closely linked to colorectal cancer due to its potential involvement in tumor progression and resistance to chemotherapy. The mechanism by which it transforms from a commensal to a pathogen remains unknown. For this reason, we investigated the role of oxidative status as an initiatory factor in changing the bacterium’s pathogenicity profile. Methods: A clinical strain of <i>F. nucleatum</i> subsp. <i>animalis</i> biofilm was exposed to different oxidative stress levels through varying subinhibitory amounts of H₂O₂. Subsequently, we investigated the bacterium’s behavior in vitro by infecting the HT-29 cell line. We evaluated bacterial colonization, volatile sulfur compounds production, and the infected cell’s oxidative status by analyzing <i>HMOX1</i>, <i>pri-miRNA 155</i>, and <i>146a</i> gene expression. Results: The bacterial colonization rate, dimethyl sulfide production, and <i>pri-miRNA 155</i> levels all increased when stressed bacteria were used, suggesting a predominant pathogenic function of these strains. Conclusions: The response of <i>F. nucleatum</i> to different oxidative conditions could potentially explain the increase in its pathogenic traits and the existence of environmental factors that may trigger the bacterium’s pathogenicity and virulence.

Jin Dai, Yifan Zhang, Jun Su et al.

BackgroundTuberculosis-associated obstructive pulmonary disease (TOPD), recognized as a high-morbidity respiratory condition in most countries, presents significant clinical challenges in differential diagnosis and comorbidity management, while substantially elevating all-cause mortality risk. Preclinical investigations of Yiqi Gubiao Pill (National Patent ZL201410536529.5) have demonstrated multi-target therapeutic efficacy, including cough suppression, bronchospasm alleviation, sputum expectoration facilitation, and disease progression retardation. This randomized controlled trial aims to systematically assess Yiqi Gubiao’s therapeutic and evaluate its safety.MethodsWe will implement a prospective double-blind randomized controlled trial utilizing 1:1 allocation ratio, randomly assigned to either the Yiqi Gubiao pill treatment group or placebo-controlled group. Following randomization, a standardized 12-week therapeutic protocol will be administered, during which serial pulmonary function assessments and quality of life evaluations will be conducted. Concurrently, validated Traditional Chinese Medicine (TCM) symptom scoring scales will be applied for score. Systematic safety surveillance will be performed through weekly monitoring of adverse events.DiscussionThis prospective, double-blind, randomized clinical trial will provide valuable data on the efficacy and safety of Yiqi Gubiao pill in treating TOPD. Positive results will offer a new treatment option for patients with TOPD.Clinical Trial Registration[ClinicalTrials.gov], identifier [NCT06676800]. Registered 30 October 2024, https://ClinicalTrials.gov.

Sébastien Dam, Jean-Marie Batail, Gabriel H Robert et al.

Background: The treatment of depressive episodes is well established, with clearly demonstrated effectiveness of antidepressants and psychotherapies. However, more than one-third of depressed patients do not respond to treatment. Identifying the brain structural basis of treatment-resistant depression could prevent useless pharmacological prescriptions,adverse events, and lost therapeutic opportunities.Methods: Using diffusion magnetic resonance imaging, we performed structural connectivity analyses on a cohort of 154 patients with mood disorder (MD) -- and 77 sex- and age-matched healthy control (HC) participants. To assess illness improvement, the MD patients went through two clinical interviews at baseline and at 6-month follow-up and were classified based on the Clinical Global Impression-Improvement score into improved or not-improved. First, the threshold-free network-based statistics was conducted to measure the differences in regional network architecture. Second, nonparametric permutations tests were performed on topological metrics based on graph theory to examine differences in connectome organization. Results: The threshold-free network-based statistics revealed impaired connections involvingregions of the basal ganglia in MD patients compared to HC. Significant increase of local efficiency and clustering coefficient was found in the lingual gyrus, insula and amygdala in the MD group. Compared with the not-improved, the improved displayed significantly reduced network integration and segregation, predominately in the default-mode regions, including the precuneus, middle temporal lobe and rostral anterior cingulate.Conclusions: This study highlights the involvement of regions belonging to the basal ganglia, the fronto-limbic network and the default mode network, leading to a better understanding of MD disease and its unfavorable outcome.

Dianna Vidas, Romina Carrasco, Ryan M. Kelly et al.

Music is a valuable non-pharmacological tool that provides benefits for people with dementia, and there is interest in designing technologies to support music use in dementia care. To ensure music technologies are appropriately designed for supporting caregivers and people living with dementia, there remains a need to better understand how music is currently used in everyday care at home. We aimed to understand how people with dementia and their caregivers use music technologies in everyday caring, as well as challenges they experience using music and technology. This study used a mixed methods design. A survey was completed by 77 caregivers and people with dementia to understand their use of music and technology. Of these, 18 survey respondents (12 family caregivers, 6 people living with dementia) participated in focus groups about their experiences of using music and technology in care. Transcripts were analysed with reflexive thematic analysis. Most survey respondents used music often in their daily lives, reporting a range of music technologies such as CDs, radio, and streaming. Focus groups highlighted benefits and challenges of music technologies in everyday care. Participants used music and music technologies to regulate mood, provide joy, facilitate social connection, encourage reminiscence, provide continuity before and after diagnosis, and to make caregiving easier. Challenges of using music technology in care included difficulties staying up to date with evolving technology, and low self-efficacy for technology use expressed by people living with dementia. Evidently, people living with dementia and their caregivers use music technologies to support their everyday care needs. Results suggest opportunities to design technologies enabling easier access to music and supporting people living with dementia with recreational and therapeutic music listening and music-based activities.

Marcin Rządeczka, Anna Sterna, Julia Stolińska et al.

Background: The increasing deployment of Conversational Artificial Intelligence (CAI) in mental health interventions necessitates an evaluation of their efficacy in rectifying cognitive biases and recognizing affect in human-AI interactions. These biases, including theory of mind and autonomy biases, can exacerbate mental health conditions such as depression and anxiety. Objective: This study aimed to assess the effectiveness of therapeutic chatbots (Wysa, Youper) versus general-purpose language models (GPT-3.5, GPT-4, Gemini Pro) in identifying and rectifying cognitive biases and recognizing affect in user interactions. Methods: The study employed virtual case scenarios simulating typical user-bot interactions. Cognitive biases assessed included theory of mind biases (anthropomorphism, overtrust, attribution) and autonomy biases (illusion of control, fundamental attribution error, just-world hypothesis). Responses were evaluated on accuracy, therapeutic quality, and adherence to Cognitive Behavioral Therapy (CBT) principles, using an ordinal scale. The evaluation involved double review by cognitive scientists and a clinical psychologist. Results: The study revealed that general-purpose chatbots outperformed therapeutic chatbots in rectifying cognitive biases, particularly in overtrust bias, fundamental attribution error, and just-world hypothesis. GPT-4 achieved the highest scores across all biases, while therapeutic bots like Wysa scored the lowest. Affect recognition showed similar trends, with general-purpose bots outperforming therapeutic bots in four out of six biases. However, the results highlight the need for further refinement of therapeutic chatbots to enhance their efficacy and ensure safe, effective use in digital mental health interventions. Future research should focus on improving affective response and addressing ethical considerations in AI-based therapy.

Radheesh Sharma Meda, Amir Barati Farimani

Identifying drug-target interactions is essential for developing effective therapeutics. Binding affinity quantifies these interactions, and traditional approaches rely on computationally intensive 3D structural data. In contrast, language models can efficiently process sequential data, offering an alternative approach to molecular representation. In the current study, we introduce BAPULM, an innovative sequence-based framework that leverages the chemical latent representations of proteins via ProtT5-XL-U50 and ligands through MolFormer, eliminating reliance on complex 3D configurations. Our approach was validated extensively on benchmark datasets, achieving scoring power (R) values of 0.925 $\pm$ 0.043, 0.914 $\pm$ 0.004, and 0.8132 $\pm$ 0.001 on benchmark1k2101, Test2016_290, and CSAR-HiQ_36, respectively. These findings indicate the robustness and accuracy of BAPULM across diverse datasets and underscore the potential of sequence-based models in-silico drug discovery, offering a scalable alternative to 3D-centric methods for screening potential ligands.

Álvaro Moreno, Lorenzo Cavicchi, Xia Wang et al.

Selective sensing of chiral molecules is a key aspect in fields spanning biology, chemistry, and pharmacology. However, conventional optical methods, such as circular dichroism (CD), encounter limitations owing to weak chiral light-matter interactions. Several strategies have been investigated to enhance CD or circularly polarised luminescence (CPL), including superchiral light, plasmonic nanoresonators and dielectric nanostructures. However, a compromise between spatial uniformity and high sensitivity, without requiring specific molecular functionalization, remains a challenge. In this work, we propose a novel approach using twisted bilayer graphene (TBG), a chiral 2D material with a strong CD peak which energy is tunable through the twist angle. By matching the CD resonance of TBG with the optical transition energy of the molecule, we achieve a decay rate enhancement mediated by resonant energy transfer that depends on the electric-magnetic interaction, that is, on the chirality of both the molecules and TBG. This leads to an enantioselective quenching of the molecule fluorescence, allowing to retrieve the molecule chirality from time-resolved photoluminescence measurements. This method demonstrates high sensitivity down to single layer of molecules, with the potential to achieve the ultimate goal of single-molecule chirality sensing, while preserving the spatial uniformity and integrability of 2D heterostructures.

Hassan N. Al Hashem, Amanda N. Abraham, Deepak Sharma et al.

The ability to form diamond electrodes on insulating polycrystalline diamond substrates using single-step laser patterning, and the use of the electrodes as a substrate that supports the adhesion and proliferation of human mesenchymal stem cells (hMSCs) is demonstrated. Laser induced graphitisation results in a conductive amorphous carbon surface, rich in oxygen and nitrogen terminations. This results in an electrode with a high specific capacitance of 182 uF/cm2, a wide water window of 3.25 V, and a low electrochemical impedance of 129 Ohms/cm2 at 1 kHz. The electrodes surface exhibited a good level of biocompatibility with hMSCs, supporting cell adhesion and proliferation. The cells cultured on the electrode displayed significant elongation and alignment along the direction of the laser patterned microgrooves. Because of its favourable electrochemical performance and biocompatibility, the laser-patterned diamond electrodes create a potential for a versatile platform in stem cell therapeutics.

Pratiksing Rajput, Kazi S. Nahar, Khondaker Miraz Rahman

The prevalence of resistance in Gram-positive bacterial infections is rapidly rising, presenting a pressing global challenge for both healthcare systems and economies. The WHO categorizes these bacteria into critical, high, and medium priority groups based on the urgency for developing new antibiotics. While the first priority pathogen list was issued in 2017, the 2024 list remains largely unchanged. Despite six years having passed, the progress that has been made in developing novel treatment approaches remains insufficient, allowing antimicrobial resistance to persist and worsen on a global scale. Various strategies have been implemented to address this growing threat by targeting specific resistance mechanisms. This review evaluates antimicrobial resistance (AMR) in Gram-positive bacteria, highlighting its critical impact on global health due to the rise of multidrug-resistant pathogens. It focuses on the unique cell wall structure of Gram-positive bacteria, which influences their identification and susceptibility to antibiotics. The review explores the mechanisms of AMR, including enzymatic inactivation, modification of drug targets, limiting drug uptake, and increased drug efflux. It also examines the resistance strategies employed by high-priority Gram-positive pathogens such as <i>Staphylococcus aureus</i>, <i>Streptococcus pneumoniae</i>, and <i>Enterococcus faecium</i>, as identified in the WHO’s 2024 priority list.

RuiXuan Liu, YunTian Luo, JinYing Ma et al.

Functional gastrointestinal disorders (FGIDs) and inflammatory bowel disease (IBD) are common clinical disorders characterized by recurrent diarrhea and abdominal pain. Although their pathogenesis has not been fully clarified, disruptions in intestinal motility and immune function are widely accepted as contributing factors to both conditions, and the brain–gut axis plays a key role in these processes. Traditional Chinese Medicine (TCM) employs a holistic approach to treatment, considers spleen and stomach impairments and liver abnormality the main pathogenesis of these two diseases, and offers a unique therapeutic strategy that targets these interconnected pathways. Clinical evidence shows the great potential of TCM in treating FGIDs and IBD. This study presents a systematic description of the pathological mechanisms of FGIDs and IBD in the context of the brain–gut axis, discusses clinical and preclinical studies on TCM and acupuncture for the treatment of these diseases, and summarizes TCM targets and pathways for the treatment of FGIDs and IBD, integrating ancient wisdom with contemporary biomedical insights. The alleviating effects of TCM on FGID and IBD symptoms are mainly mediated through the modulation of intestinal immunity and inflammation, sensory transmission, neuroendocrine–immune network, and microbiota and their metabolism through brain–gut axis mechanisms. TCM may be a promising treatment option in controlling FGIDs and IBD; however, further high-quality research is required. This review provides a reference for an in-depth exploration of the interventional effects and mechanisms of TCM in FGIDs and IBD, underscoring TCM’s potential to recalibrate the dysregulated brain–gut axis in FGIDs and IBD.

Boitumelo Setlhare, Marothi Letsoalo, Siphathimandla Authority Nkabinde et al.

Alann Caderhoussin, David Couvin, Gaëlle Gruel et al.

IntroductionThis study aimed to understand the origin and to explain the maintenance of extended-spectrum β-lactamase (ESBL) Enterobacteriaceae isolated from food-producing animals in a third-generation cephalosporin (3GC)-free farm.MethodsCulture and molecular approaches were used to test molecules other than 3GC such as antibiotics (tetracycline and oxytetracycline), antiparasitics (ivermectin, flumethrin, fenbendazol, and amitraz), heavy metal [arsenic, HNO3, aluminum, HNO3, cadmium (CdSO4), zinc (ZnCl2), copper (CuSO4), iron (FeCl3), and aluminum (Al2SO4)], and antioxidant (butylated hydroxytoluene) as sources of selective pressure. Whole-genome sequencing using short read (Illumina™) and long read (Nanopore™) technologies was performed on 34 genomes. In silico gene screening and comparative analyses were used to characterize the genetic determinants of resistance, their mobility, and the genomic relatedness among isolates.ResultsOur analysis unveiled a low diversity among the animal ESBL-producing strains. Notably, E. coli ST3268 was recurrently isolated from both flies (n = 9) and cattle (n = 5). These E. coli ST3268/blaCTX-M-15/blaTEM-1B have accumulated multiple plasmids and genes, thereby representing a reservoir of resistance and virulence factors. Our findings suggest that flies could act as effective mechanical vectors for antimicrobial gene transfer and are capable of transporting resistant bacteria across different environments and to multiple hosts, facilitating the spread of pathogenic traits. A significantly higher mean minimum inhibitory concentration of oxytetracycline (841.4 ± 323.5 mg/L vs. 36.0 ± 52.6 mg/L, p = 0.0022) in ESBL E. coli than in non-ESBL E. coli and blaCTX-M-15 gene overexpression in oxytetracycline-treated vs. untreated ESBL E. coli (RQOxy = 3.593, p = 0.024) confirmed oxytetracycline as a source of selective pressure in ESBL E. coli.DiscussionThe occurrence of ESBL E. coli in a farm without 3GC use is probably due to an as yet undefined human origin of Enterobacteriaceae blaCTX-M-15 gene transmission to animals in close contact with cattle farm workers and the maintenance of the local ESBL E. coli reservoir by a high fly diversity and oxytetracycline selective pressure. These findings highlight the critical need for stringent vector control to mitigate antimicrobial resistance spread for preserving public health. Addressing this issue necessitates a multifaceted approach combining microbial genetics, vector ecology, and farm management practices.

Sabrina Haque, Akash Ahmed, Nazrul Islam et al.

Recent research has shown that antibiotic-resistant microorganisms are becoming more prevalent in intensive care units (ICUs) at an exponential rate. Patients in the ICU can get infected by pathogens due to invasive operation procedures and critical health conditions. This study primarily emphasized tracheal samples from ICU patients due to their reliance on ventilators, increasing their susceptibility to Ventilator-Associated Pneumonia (VAP). Moreover, the rise of multidrug-resistant (MDR) pathogens makes treatment strategies more challenging for these patients. In this study, we tested 200 tracheal specimens to determine the prevalence of microorganisms and analyzed the antibiotic susceptibility of these isolates against regular antibiotics, including 4th generation drugs. Among the 273 isolates, 81% were gram-negative bacteria, 10% were gram-positive bacteria, and 9% were fungi. The most prevalent gram-negative bacteria were <i>Acinetobacter</i> spp. (34%), <i>Klebsiella</i> spp. (22%), <i>Pseudomonas</i> spp. (14%), and <i>Escherichia coli</i> (9.2%). The most prevalent gram-positive bacteria were <i>Staphylococcus aureus</i> (5.9%), and the fungi were <i>Candida</i> spp. (7.3%). Among the most prevalent bacteria, except <i>Staphylococcus aureus</i> isolates, around 90% were resistant to multiple drugs, whereas 60% of <i>Acinetobacter</i> spp. and <i>Pseudomonas</i> spp. were extensively drug resistant. Sensitivity analysis against the gram-negative and gram-positive drug panel using a one-way ANOVA test followed by Tukey’s post hoc test showed that in the in vitro assay, colistin was the most effective antibiotic against all gram-negative bacteria. In contrast, linezolid, vancomycin, and fusidic acid were most effective against all gram-positive bacteria. Regular monitoring of nosocomial infections and safe management of highly resistant bacteria can help prevent future pandemics.

Ying-Chao Liang, Ling Li, Jia-Lin Liang et al.

BackgroundAlternative and complementary therapies play an imperative role in the clinical management of Type 2 diabetes mellitus (T2DM), and exploring and utilizing natural products from a genetic perspective may yield novel insights into the mechanisms and interventions of the disorder.MethodsTo identify the therapeutic target of baicalin for T2DM, we conducted a Mendelian randomization study. Druggable targets of baicalin were obtained by integrating multiple databases, and target-associated cis-expression quantitative trait loci (cis-eQTL) originated from the eQTLGen consortium. Summary statistics for T2DM were derived from two independent genome-wide association studies available through the DIAGRAM Consortium (74,124 cases vs. 824,006 controls) and the FinnGen R9 repository (9,978 cases vs. 12,348 controls). Network construction and enrichment analysis were applied to the therapeutic targets of baicalin. Colocalization analysis was utilized to assess the potential for the therapeutic targets and T2DM to share causative genetic variations. Molecular docking was performed to validate the potency of baicalin. Single-cell RNA sequencing was employed to seek evidence of therapeutic targets’ involvement in islet function.ResultsEight baicalin-related targets proved to be significant in the discovery and validation cohorts. Genetic evidence indicated the expression of ANPEP, BECN1, HNF1A, and ST6GAL1 increased the risk of T2DM, and the expression of PGF, RXRA, SREBF1, and USP7 decreased the risk of T2DM. In particular, SREBF1 has significant interaction properties with other therapeutic targets and is supported by strong colocalization. Baicalin had favorable combination activity with eight therapeutic targets. The expression patterns of the therapeutic targets were characterized in cellular clusters of pancreatic tissues that exhibited a pseudo-temporal dependence on islet cell formation and development.ConclusionThis study identified eight potential targets of baicalin for treating T2DM from a genetic perspective, contributing an innovative analytical framework for the development of natural products. We have offered fresh insights into the connections between therapeutic targets and islet cells. Further, fundamental experiments and clinical research are warranted to delve deeper into the molecular mechanisms of T2DM.

Muhtar Kadirhaz, Yushan Zhang, Naveel Atif et al.

Objectives: Overuse of antibiotics in healthcare remains prevalent and requires urgent attention in China, particularly in primary healthcare (PHC) facilities. This study aimed to describe the patterns of antibiotic prescriptions for acute upper respiratory tract infections (URTIs) in PHC facilities in China and to investigate how PHC physicians’ knowledge influences their antibiotic prescribing behavior. Methods: A cross-sectional survey linking physician questionnaire responses and their prescription data was conducted in Shaanxi Province, China. The proportions of URTI visits that received at least one antibiotic, combined antibiotics, and broad-spectrum antibiotics were the main outcomes reflecting antibiotic prescribing behavior. Multivariate mixed-effects logistic regressions were applied to analyze the relationship between PHC physicians’ knowledge about antibiotics and their antibiotic prescribing behavior. Results: A total of 108 physicians filled out the questionnaires between February 2021 and July 2021, and a sample of 11,217 URTI visits attended by these physicians from 1 January 2020 to 31 December 2020 were included in the analysis. The overall mean score of our respondents on the knowledge questions was 5.2 (total score of 10). Over sixty percent (61.2%; IQR 50.2–72.1) of the URTI visits received antibiotics. The percentages of URTI visits prescribed with combined and broad-spectrum antibiotics were 7.8% (IQR 2.3–10.2) and 48.3% (IQR 36.7–58.7), respectively. Third-generation cephalosporins were the most frequently used antibiotics. Physicians with lower antibiotic knowledge scores were more inclined to prescribe antibiotics (<i>p</i> < 0.001), combined antibiotics (<i>p</i> = 0.001), and broad-spectrum antibiotics (<i>p</i> < 0.001). Conclusions: Physicians’ insufficient knowledge was significantly associated with antibiotic overprescribing. Professional training targeting physicians’ knowledge of antibiotics is urgently needed to improve the rational use of antibiotics in grassroots healthcare facilities in China.

Hongxin Li, Jingsheng Tang, Wenqi Li et al.

Electroencephalography (EEG) datasets are characterized by low signal-to-noise signals and unquantifiable noisy labels, which hinder the classification performance in rapid serial visual presentation (RSVP) tasks. Previous approaches primarily relied on supervised learning (SL), which may result in overfitting and reduced generalization performance. In this paper, we propose a novel multi-task collaborative network (MTCN) that integrates both SL and self-supervised learning (SSL) to extract more generalized EEG representations. The original SL task, i.e., the RSVP EEG classification task, is used to capture initial representations and establish classification thresholds for targets and non-targets. Two SSL tasks, including the masked temporal/spatial recognition task, are designed to enhance temporal dynamics extraction and capture the inherent spatial relationships among brain regions, respectively. The MTCN simultaneously learns from multiple tasks to derive a comprehensive representation that captures the essence of all tasks, thus mitigating the risk of overfitting and enhancing generalization performance. Moreover, to facilitate collaboration between SL and SSL, MTCN explicitly decomposes features into task-specific features and task-shared features, leveraging both label information with SL and feature information with SSL. Experiments conducted on THU, CAS, and GIST datasets illustrate the significant advantages of learning more generalized features in RSVP tasks. Our code is publicly accessible at <uri>https://github.com/Tammie-Li/MTCN</uri>.