搜索结果：Frames

Object detection in soccer videos plays an important role in intelligent broadcasting, tactical analysis, and player tracking. Frequent view switching driven by multi-camera coordination in soccer broadcasts introduces two distinct challenges for existing detection methods. First, in the temporal dimension, shot transitions cause abrupt visual changes between adjacent frames. Existing temporal aggregation methods forcibly fuse features from heterogeneous views, injecting cross-view noise that can degrade detection accuracy below single-frame baselines. Second, in the feature dimension, target appearance varies significantly across views: players in panoramic shots are low-resolution small objects characterized mainly by coarse contours and color, whereas the same players in close-up shots become high-resolution large objects with rich texture and fine-grained detail. A unified detection strategy struggles to handle both extremes simultaneously. We propose ViewAdapt-Det, a framework consisting of two complementary modules. The View-Aware Temporal Gating module (VATG) detects shot transitions and dynamically controls temporal aggregation intensity, suppressing historical frame contributions during view changes while preserving temporal enhancement during continuous sequences. The View-Conditioned Detection Modulation module (VCDM) infers the current view type and conditionally modulates detection features, allowing the detector to apply view-specific feature processing strategies that match the visual characteristics of each viewpoint. Experiments on SoccerNet-Tracking, SportsMOT, and our self-constructed BroadcastSwitch-Soccer dataset show that ViewAdapt-Det outperforms existing methods and demonstrates stronger robustness under abrupt shot transitions.

Advances in three-dimensional magnetic resonance spectroscopic imaging (3D-MRSI) allow for the high-resolution mapping of multiple neurometabolites throughout the entire brain in vivo and within clinically compatible time frames. Leveraging this capability, we created a voxel-based pipeline that corrects and spatially normalizes whole-brain maps of total N-acetylaspartate (tNAA), myo-inositol (Ins), choline compounds (Cho), glutamate + glutamine (Glx) and creatine + phosphocreatine (tCr). We examined 2 different 3D-MRSI dataset: first, a clinical sample of adolescents and young adults at risk for psychosis (n = 21) meeting DSM-5 criteria for Attenuated Psychosis Syndrome (APS) or Schizotypal Personality Disorder (SCZT), and age-/sex-matched healthy controls (n = 13); and second, a non-clinical sample of adolescents (n = 61) scanned on a different site. The objective of the study was threefold: first, to assess the reproducibility of 3D-MRSI measures across datasets and scanning sites; second, to validate the feasibility of whole-brain, voxel-based analyses on 3D-MRSI data; and third, to test the sensitivity of this approach. Metabolite distributions showed reproducible regional variation in standard space between the two independent samples and scanning sites (r ranging from 0.82 to 0.99). Relative to controls, at-risk participants exhibited higher tNAA levels in frontal grey matter; the SCZT subgroup additionally displayed widespread cortical and subcortical elevations of Ins levels compared with both APS and controls. Voxel-based analyses of structural (i.e., gray and white matter volumes or densities) and diffusion (i.e., generalized fractional anisotropy) parameters yielded no significant differences between patients and controls. These preliminary findings suggest that high-resolution 3D-MRSI may be sensitive enough to detect subtle neurometabolic alterations at the group level in the early stages of psychotic disorders when structural or diffusion measures show no difference. High-resolution whole-brain metabolic mapping may have the potential to help with early identification of young people at risk for psychosis or other mental disorders.

This article presents a collection of 360 synchronized stereo video pairs capturing the flight behaviour of Budgerigars (Melopsittacus undulatus) in a controlled indoor arena. The recordings were acquired at 120 frames per second using a calibrated stereo camera setup with a fixed baseline. The dataset includes both solo flight sequences and structured head-on interaction scenarios involving one-to-one, two-to-two, and three-to-three group configurations. All individuals are manually annotated in both camera views using identity-consistent bounding boxes along with four body keypoints: head, tail, left wing, and right wing. In total, the dataset contains 2760,178 labelled annotations. The synchronized two-dimensional annotations from both views are reconstructed into metric three-dimensional coordinates using calibrated stereo triangulation, providing frame-level trajectories and pose information. The dataset includes raw stereo video files, annotation files in CVAT XML format, processed three-dimensional trajectory data in CSV format, stereo calibration parameters, and scripts for reconstruction. Technical validation measures are provided to document dataset quality, including stereo calibration consistency assessed using stereo reprojection error of 0.50 pixels, annotation reliability analysis, and qualitative assessment of physically plausible motion patterns. By integrating synchronized stereo recordings, identity-consistent 2D annotations, and reconstructed 3D trajectories within a single resource, the dataset supports applications such as multi-object tracking, 3D pose estimation, trajectory modelling, and the analysis of multi-agent interaction, while detailed wing kinematic analysis is constrained by interpolation uncertainty and rolling-shutter effects.

Ferroptosis is an iron-dependent form of regulated cell death driven by phospholipid peroxidation. In the central nervous system (CNS), most ferroptosis research has focused on neurons and glial cells, whereas the vulnerability of brain microvascular endothelial cells (BMECs) and its consequences for blood-brain barrier (BBB) integrity remain less clearly defined. Because BMECs form the vascular interface between the circulation and the brain parenchyma, ferroptotic injury in this cell population may represent an immunovascular mechanism through which endothelial redox stress is translated into barrier dysfunction and neuroinflammatory amplification. In this review, we summarize molecular pathways that may promote or restrain BMEC ferroptosis, including iron handling, antioxidant defense mediated by the solute carrier family 7 member 11 (SLC7A11)-glutathione peroxidase 4 (GPX4) axis and nuclear factor erythroid 2-related factor 2 (Nrf2) signaling, lipid peroxidation, and junctional remodeling. We then discuss how ferroptosis-associated endothelial injury may contribute to BBB leakage, damage-associated molecular pattern release, innate immune sensing, leukocyte recruitment, glial activation, and self-amplifying inflammatory feedback at the neurovascular interface. We organize the available literature according to the strength and cellular specificity of evidence, separating BMEC-specific findings, BBB-focused in vivo studies, indirect CNS evidence, and mechanistic analogies from non-CNS endothelial systems. Finally, we evaluate disease-specific evidence in ischemic stroke and selected neurodegenerative or inflammatory conditions, together with therapeutic strategies, BMEC-targeting considerations, candidate clinical biomarkers, and translational barriers for modulating endothelial ferroptosis. This review frames endothelial ferroptosis as a promising but incompletely established immunovascular link between BBB dysfunction and neuroinflammation, and highlights the need for BMEC-specific models, human BBB systems, endothelial ferroptosis biomarkers, biomarker-guided monitoring, BMEC-targeted delivery approaches, and careful evaluation of the physiological risks of systemic or prolonged ferroptosis blockade.

Understanding cellular dynamics represents a critical challenge in biomedical research. Optical microscopy remains a key technique for observing live-cell behaviors in vitro. This paper introduces an enhanced cell-tracking algorithm designed to address dynamic changes in cell populations, including mitosis, migration, and cell-cell interactions, even within complex co-culture models. The proposed method involves three main steps: 1) modeling the movements and interactions of different cell types in co-culture experiments via tailored open multi-agent systems; 2) identifying parameters via real data for a multi-agent, multi-culture framework; 3) embedding the model within an Extended Kalman Filter, to predict the dynamics of heterogeneous cell populations across video frames. To validate the approach, we used a novel dataset involving the interplay between tumor and normal cells, namely osteosarcoma and mesenchymal stromal cells, respectively. This dataset offers a challenging and clinically relevant framework to track cell proliferation and study how cancer cells evolve and interact with stromal cells within their surroundings. Performance metrics demonstrated the effectiveness of the algorithm over state-of-the-art methodologies, highlighting its ability to track heterogeneous cell types, capture their interactions, and generate the estimated cell lineage tree.

Policy Points Chronic absence should be recognized as a public health indicator and early warning sign that systems are failing to meet the developmental, social, and health needs of students. Improving student attendance requires cross-sector policy action across education, health, and public health to address the structural and social determinants of chronic absence. A prevention-oriented public health approach is essential, focusing on root causes that schools cannot address alone such as poor health, housing instability, and unreliable transportation. Chronic absence, defined as missing more than 10% of time in school, has risen sharply in the United States following the COVID-19 pandemic and now affects more than one in four students. It reflects unmet health and social needs and is patterned by deep structural inequalities. Both short- and long-term consequences include adverse impacts on educational attainment, health, and social outcomes. Despite this, chronic absence remains largely framed and addressed as an education-sector problem, limiting the scope and effectiveness of current responses. This perspective synthesizes interdisciplinary evidence from education, public health, and child development literature, drawing on ecological and life course frameworks to reconceptualize chronic absence as a public health issue. We develop a conceptual model integrating multilevel determinants of attendance across individual, family, school, community, and structural domains, and identify implications for policy and cross-sector action. Viewing chronic absence through a public health lens reframes it from a purely educational outcome to a signal of unmet need and a multidimensional indicator of system performance. Attendance patterns reflect the interaction of health, social, and structural factors that lie largely outside of the control of schools. Current approaches often emphasize individual responsibility, while overlooking the broader conditions that shape attendance. Reframing chronic absence in this way underscores the need for coordinated cross-sector interventions that address underlying determinants. Positioning chronic absence as a public health priority enables a more coherent response. We propose three principles to guide action: (1) use school attendance data as a vital sign of student and system well-being; (2) develop strategic partnerships to align goals and drive progress; and (3) develop strengths-based policies and programs to prevent chronic absence. Without this shift, efforts to reduce chronic absence are likely to remain fragmented and insufficient to achieve equitable improvements in child health and educational outcomes.

To evaluate the first audiometric outcomes of Nuance Audio hearing glasses, an over-the-counter solution integrating air conduction amplification technology into standard eyeglass frames, in individuals with mild to moderate hearing loss. Thirty-two adults (mean age 74 years) with symmetric, age-related mild to moderate sensorineural hearing loss were tested. Pure-tone and speech audiometry were performed under unaided and aided free-field conditions. Device settings were adjusted via smartphone app based on audiometric profiles and user preference. Paired t-tests assessed the differences. Nuance Audio glasses significantly improved hearing thresholds, especially at 4000-6000 Hz, with mean gains of 10-11 dB HL. Speech reception thresholds improved by 7 dB at 50% intelligibility and 6 dB at 100%. Amplification profile preferences matched audiometric slopes and no differences in benefit were found between mild and moderate hearing loss groups. Nuance Audio glasses provide clinically relevant improvements in pure-tone and speech thresholds for adults with mild to moderate presbycusis. Their discreet, open-ear design and self-fitting approach address barriers to traditional hearing aids and may support earlier adoption.

Estimating vocal-tract length (VTL) from vowel formants can aid speaker normalization, but few methods have been benchmarked against an anatomical reference in the same speakers. We combined acoustic pharyngometry (APh) and speech data from 42 adults to benchmark eight widely used formant-based VTL estimators against incisors-to-glottis length and to test an interpretable two-stage bias-corrected linear estimator. Across more than 400 000 central frames with valid F1-F4, traditional quarter-wave, odd-harmonic, and dispersion-type estimators correlated with VTLAPh but showed poor out-of-sample anatomical recovery and strong calibration compression. Re-estimated one-stage linear models reduced mean absolute error (MAE; median ≈1.0 cm) but still overestimated shorter tracts and underestimated longer tracts. A two-stage model markedly improved calibration and agreement, outperforming one-stage linear and nonlinear alternatives (median per-vowel MAE 0.39 cm, median out-of-sample R2=0.83). Front and front-rounded vowels were the most informative. Speaker-level 95% limits of agreement were about ±0.9 cm, indicating that the method is better suited to aggregated tract-scale estimation than to direct anatomical measurement. These results identify calibration bias as a central limitation of standard formant-based VTL estimators and provide a practical, interpretable route to tract-scale estimation from similarly processed labeled-vowel data under matched conditions.

To evaluate the feasibility of using convolutional neural networks (CNNs) and vision transformers (ViTs) to predict renal tumor pathology intraoperatively based on gross appearance. Intraoperative images were retrospectively extracted from surgical recordings of patients undergoing partial nephrectomy between 2008-2024. Static frames obtained prior to arterial clamping were curated and linked with final pathology. A ResNet50-based CNN and the General Surgery Vision Transformer (GSViT) were trained to classify six tumor types: clear cell RCC (ccRCC), papillary RCC (pRCC), chromophobe RCC (chRCC), hybrid oncocytic tumors, oncocytoma, and angiomyolipoma (AML). Models were trained with transfer learning, evaluated on held-out test data, and assessed using accuracy, AUC-ROC, and confusion matrices. A total of 443 images from 118 patients (136 surgeries) were analyzed, including ccRCC (n=149), pRCC (n=97), chRCC (n=42), hybrid tumors (n=81), oncocytoma (n=43), and AML (n=31). In binary classification, the CNN achieved the highest AUCs for ccRCC (0.74), chRCC (0.70), hybrid tumors (0.73), and AML (0.70). Multi-class CNN performance was more variable, with notable AUCs for pRCC (0.70) and oncocytoma (0.71). The GSViT model underperformed across most categories, demonstrating prediction bias toward ccRCC. Attempts to unfreeze pretrained backbones led to rapid overfitting, underscoring dataset limitations. CNN-based models demonstrate moderate ability to classify renal tumor pathology intraoperatively from gross appearance, providing proof of concept for AI-assisted surgical decision-making. Larger datasets and external validation are needed before clinical application.

Classical pattern-triggered (PTI) and effector-triggered (ETI) immunity, developed in single-pathogen systems, illuminates how plants recognise molecular threats but cannot fully explain immune homeostasis within the dynamic microbial communities plants encounter in nature. The extended plant immune system reframes immunity as a host-microbiome network sculpted by root exudates, yet two dimensions remain insufficiently integrated: the ecological rules translating recruited communities into systemic immune output, and the mechanisms by which holobiont state may carry over across generations. We propose that plant immune homeostasis is best analyzed as a three-node feedback circuit that we hypothesize closes across generations. Node 1 (molecular recruitment) integrates root exudate-mediated cross-kingdom signalling, in which primary and secondary metabolites jointly serve nutritional and immune-informative roles. Node 2 (ecological translation) is governed by dispersal, immune filtering, drift, priority effects, and functional redundancy, which together determine whether recruitment signals translate into immune buffering. Node 3 (intergenerational carry-over) comprises three mechanistically distinct routes-epigenetic reprogramming, seed microbiota transmission, and soil legacy-that range from provisionally established to largely hypothetical and whose field-scale validation remains limited. Treating this circuit, rather than the host or host-microbiome network, as the minimal unit of immune analysis generates testable predictions-linking functional redundancy to immune buffering, soil legacy to next-generation priming, and node-specific failure modes to dissociable signatures. This framing positions the holobiont across time (understood here as an analytical unit rather than an evolutionary one) as a tractable framework for hypothesis-driven plant immunity research.

Antimicrobial resistance poses an increasing global challenge, driving the urgent need for alternative strategies to identify novel therapeutic agents. Microbial natural products encoded by biosynthetic gene clusters (BGCs) remain among the most promising sources of bioactive compounds. Although Corynebacterium glutamicum is best known as an industrial producer of amino acids, its potential as a producer of secondary metabolites has not been comprehensively assessed, despite the availability of numerous high-quality genome sequences. In this study, we carried out a comparative pangenome analysis of 36 complete C. glutamicum genomes and systematically mined for BGCs to explore the species' biosynthetic repertoire. Our analysis revealed variation in BGC content among strains, with several isolates harboring more hybrid clusters than others, suggesting metabolic diversity across the species. In addition to conserved terpene biosynthetic pathways, we detected polyketide-associated clusters not previously reported in C. glutamicum, expanding its recognized metabolic potential. RiPP-like clusters, including Lactococcin-related variants, were also identified, highlighting an underexplored reservoir of antimicrobials. To prioritize candidates for future validation, Support Vector Machine, Random Forest, and k-Nearest Neighbor models were trained on Composition, Transition, and Distribution (CTD) physicochemical sequence features and applied to genome-mined small open reading frames. The models demonstrated strong predictive performance, with the Support Vector Machine achieving the highest accuracy (84.1%), F1 score (83.8%), and area under the ROC curve (AUC = 0.920). After removing duplicate sequence IDs and applying a high-confidence AMP probability threshold (≥0.95), 18 unique AMP-like candidates were identified as promising. Overall, this study presents C. glutamicum as a promising source of bioactive metabolite candidates and shows how pangenome-scale mining combined with machine learning can support antimicrobial peptide discovery while still requiring experimental validation of the predicted leads.

Automated segmentation of breast lesions in ultrasound videos is critical for clinical applications but remains hindered by the reliance on expensive pixel-wise annotations. While scribble supervision offers a user-friendly alternative, its potential for ultrasound video segmentation remains underexplored. In this regard, we propose ScribSAM, a novel scribble-supervised framework built on the Medical Segment Anything Model (MedSAM) for robust ultrasound video segmentation. ScribSAM integrates two key innovations: a flow-guided scribble propagation module that leverages optical flow to efficiently propagate sparse scribble annotations across frames while preserving temporal consistency, and a bidirectional cross-attention module that fuses MedSAM's global ViT embeddings with 3D CNN local-temporal embeddings for enhanced spatiotemporal feature learning. Extensive experiments on the scribble-annotated variants of two ultrasound video datasets, BUV2022 and US-VOS, demonstrate ScribSAM's superiority. It surpasses state-of-the-art scribble-supervised methods by 5.53% (BUV2022) and 8.31% (US-VOS) in Dice score, outperforming some fully supervised methods and substantially narrowing the gap with the best fully supervised methods, while using only 4% of the annotated pixels. Code and dataset will be released at https://github.com/003-GH/ScribSAM.

Atrial fibrillation is the most common cardiac arrhythmia in clinical practice making proficient diagnosis and treatment vital to patient outcomes. The increased incidence of new-onset symptomatic rapid atrial fibrillation (NOSRAF) necessitates consistent provider training. This practice improvement project created a simulation-based training (SBT) program to prepare prehospital health care providers for responding to NOSRAF. A pre/post design was used to evaluate SBT for prehospital health care providers during biannual training. A protocol-based checklist, knowledge questionnaire, and the Simulation Effectiveness Tool-Modified (SET-M) were used to collect data on protocol fidelity, knowledge, and simulation effectiveness, respectively. SBT improved NOSRAF protocol compliance; however, the SET-M survey data revealed no change in provider confidence. Further studies focused on SBT for the prehospital health care provider are warranted. SBT can improve prehospital health care provider preparedness in caring for patients with NOSRAF. Further studies are needed to better represent the influence of simulation-based learning in the prehospital health care provider population and to identify knowledge deterioration time frames.

The rapid expansion of telemedicine has increased the relevance of teleconsultations in everyday clinical practice. However, the manner in which teleconsultation competencies are conceptualized, taught, and assessed in undergraduate medical education remains unclear. A narrative review was conducted using a transparent methodology to identify and select studies. Several bibliographic databases were searched using predefined eligibility criteria, focusing on educational interventions that specifically targeted teleconsultation competencies among undergraduate medical students. The selection process emphasized conceptual clarity, curricular intent, and the inclusion of empirically reported educational outcomes. The eligibility criteria were intentionally designed to identify studies that explicitly conceptualized teleconsultation as a distinct undergraduate clinical competency. The search process yielded 199 records. After screening and comprehensive full-text evaluation, only one study explicitly conceptualized teleconsultation as a distinct undergraduate clinical competency and met all predefined eligibility criteria. This study details a structured educational intervention that frames teleconsultation as a distinct clinical competency. The outcomes concerning communication, clinical reasoning in virtual environments, technical aspects of teleconsultation, and components of remote physical examination were reported. The principal finding was the identification of substantial evidence gaps in the conceptualization, assessment, and longitudinal teaching of teleconsultation competencies. Current evidence regarding the instruction of teleconsultation competencies in undergraduate medical education is limited. The prevailing literature addresses telemedicine primarily as a mode of care delivery rather than emphasizing teleconsultation as a distinct clinical skill. These observations highlight a discrepancy between contemporary clinical practice and undergraduate training, emphasizing the need for well-defined educational frameworks that incorporate teleconsultation competencies.

This conceptual article proposes a patient-counseling framework that integrates epigenetic aging measures into the Transtheoretical Model (TTM) to support personalized health behavior change. We developed a conceptual model by mapping potentially motivational properties of epigenetic age feedback, including its interpretability, biological science, and modifiability, onto stage-specific behavior-change processes within the TTM. The proposed model describes how epigenetic age feedback may be introduced differently across precontemplation, contemplation, preparation, action, maintenance, and relapse phases. Rather than functioning as a stand-alone motivator, epigenetic age is positioned as a complementary tool alongside established approaches such as motivational interviewing, wearable monitoring, and conventional risk assessment. This framework also outlines how repeat measurement may be used selectively to reinforce progress, re-engage patients after slippage, and personalize counseling around behaviors such as physical activity, sleep, nutrition, and stress management. Integrating epigenetic age data into stage-matched counseling offers a plausible, patient-centered approach to behavior change, but it remains conceptual and requires empirical evaluation of feasibility, acceptability, and communication effects. This article frames epigenetic age from a research biomarker into a stage-specific patient education tool and proposes a modified TTM in which repeat biological age assessment is used as a targeted relapse-prevention and re-engagement strategy.

This article examines development financing as a field of global development governance by analyzing an academic deliberative space organized in parallel to the Fourth United Nations Conference on Financing for Development (FfD4). Drawing on critical development theory and critical perspectives on human rights, the study explores how expert debates respond to and reinterpret dominant framings of development financing as a neutral and technical set of instruments oriented toward efficient resource mobilization. Methodologically, the research combines qualitative discourse analysis with semantic network analysis, based on a systematically constructed corpus of recorded expert interventions and moderated debates produced during the academic event. The corpus includes contributions from diverse institutional actors participating in structured thematic panels. Using software-assisted coding (ATLAS.ti) and network analysis (UCINET), the study applies centrality measures (degree, closeness, and betweenness) to operationalize discourse as a relational structure. Analytical rigor was strengthened through iterative coding refinement, team-based review of categories, and structural robustness checks of the generated semantic networks. The findings reveal a hierarchically organized network in which concepts associated with cooperation, governance, sustainability, and innovation occupy structurally central positions, while narratives linked to social justice, community processes, and the expansion of rights remain comparatively peripheral. The findings suggest that development financing is predominantly framed through technocratic and procedural rationalities, while critical ethical perspectives circulate with more limited structural influence. By integrating discourse theory and social network analysis, the article contributes to sociological debates on global governance, power, and the organization of normative hierarchies in contemporary development.

This data article describes a structured dataset of seismic and geodetic deformation parameters for the Kopeh Dagh Belt in northeastern Iran. The dataset integrates Global Navigation Satellite System (GNSS) velocity measurements from 16 stations with earthquake focal mechanism information covering the period 1969-2024. GNSS velocity components and associated uncertainties are provided in a Eurasia-fixed reference frame. Earthquake data include location, magnitude, focal depth, and nodal plane parameters (strike, dip, and rake), together with derived stress tensor quantities. The spatial framework is defined by a discretization of the region into 20 triangular subnetworks using Delaunay triangulation. For each subnetwork, geodetic strain components are calculated from GNSS velocity gradients, while seismic strain parameters are derived from moment tensor information. Additional variables include rotation rates obtained from the antisymmetric component of the velocity gradient tensor, as well as principal stress axes, stress ratios, and stress regime classifications obtained through inversion of focal mechanism data. All datasets are organized in a machine-readable format and include raw measurements, intermediate parameters, and derived quantities. The dataset is intended for reuse in tectonic analyses, numerical modeling of deformation fields, evaluation of strain partitioning, and comparison of seismic and geodetic deformation patterns in intraplate regions.

Population aging raises uncertainty about future healthcare costs, pension adequacy, and financial security, yet evidence remains limited on how aging-related risk perceptions shape private medical insurance (PMI) intentions among adult consumers aged 18-64. This study examines whether perceived aging risk (AR) is associated with PMI purchase intentions and whether education, subjective health status, and national context condition this association. Survey data from consumers aged 18-64 in China (N = 484) and Malaysia (N = 415) were analyzed using PLS-SEM. Two country-specific models and one pooled boundary-condition model (N = 899) were estimated, with demographic, health, insurance-experience, and income controls. AR was consistently and positively associated with PMI purchase intentions across all models, whereas education and subjective health status showed no direct effects. In Malaysia, the AR-intention association was stronger among lower-education respondents and those reporting better subjective health; these moderation effects were not observed in China. The pooled model indicated that national context conditioned only education-based moderation. The findings suggest that AR is a robust expectancy-based correlate of PMI intentions, but its translation into intention depends on context-specific consumer resources. PMI communication in aging societies should combine aging-risk framing with locally calibrated education-sensitive implementation.

Emotions are closely implicated in how states are represented and interpreted. This study examines the affective dimensions of South Korean media portrayals of the United States and China from 1992 to 2025, focusing on variation in evaluative tone and emotional intensity. Drawing on theories of affective framing and media agenda-setting, the analysis traces long-term patterns in valence and arousal using a large corpus of newspaper articles and a computational text analysis approach. The results indicate a persistent asymmetry. Coverage of the United States remains relatively stable, with moderately positive evaluations and low levels of emotional intensity, whereas portrayals of China become more negative and more affectively charged over time. These differences appear as sustained patterns rather than isolated fluctuations and are consistent with broader features of South Korea's geopolitical context, including alliance relations, historical memory, and media environments. By documenting these patterns, this study contributes to the analysis of affect in media discourse and provides a framework for examining how major powers are represented in national contexts over time.