Search Results (6,692)

Lung cancer remains the leading cause of cancer-related mortality worldwide, emphasizing the critical need for novel and robust biomarkers to improve prognostication and guide precision oncology. While traditional clinical variables such as tumor stage, age, and sex are routinely used for survival prediction, their prognostic performance is limited. Imaging biomarkers derived from radiomic analysis of advanced medical imaging have emerged as a promising class of noninvasive cancer biomarkers, enabling quantitative characterization of tumor phenotypes. In this study, we investigated the prognostic utility of radiomic imaging biomarkers, with a particular focus on the texture-based feature Busyness, and compared their performance against conventional clinical factors. Survival analyses demonstrated that Busyness achieved significantly stronger discrimination of survival outcomes than stage, age, or sex. Stratified analyses further showed that Busyness consistently remained a dominant predictor of survival across age and sex subgroups, whereas tumor stage alone provided limited prognostic separation. To address class imbalance and enhance model robustness, the Synthetic Minority Over-sampling Technique (SMOTE) was applied, further supporting the stability of the imaging biomarker findings. These results highlight the potential of radiomic imaging biomarkers as powerful prognostic tools in lung cancer and support their integration into clinical workflows. This work contributes to the growing landscape of new cancer biomarkers and provides a foundation for future studies integrating imaging biomarkers with molecular and genomic markers to achieve improved prognostic accuracy. Full article

Purpose: Kawasaki disease (KD) is an acute systemic vasculitis that can result in coronary artery lesions (CALs). This study aims to explore the expression levels of serum Ninjurin-1 (NINJ1) and high-mobility group box 1 (HMGB1) in the acute phase of KD and evaluate their clinical significance. Methods: A total of 180 children were enrolled, comprising 113 KD patients, 35 healthy controls (HCs), and 32 febrile controls whose clinical data were collected. Serum levels of NINJ1, HMGB1, Lactate Dehydrogenase (LDH), and routine inflammatory markers were compared across groups. Serum levels of NINJ1 and HMGB1 were measured via ELISA. Correlations were analyzed using Spearman tests. The diagnostic and predictive performance of biomarkers was assessed using Receiver Operating Characteristic (ROC) curve analyses. Results: Serum levels of NINJ1 and HMGB1 were significantly elevated in the KD group compared with both the HC and FC groups (all p < 0.001). NINJ1 levels were positively correlated with the z-scores of coronary arteries and were significantly higher in the CAL subgroup than in the non-CAL subgroup (p = 0.004). A strong positive correlation was observed between serum NINJ1 and HMGB1 levels in the KD group (p < 0.001). Conclusions: Elevated serum NINJ1 levels during the acute phase of KD were associated with the presence of CALs, while HMGB1 shows promise in differentiating KD from other febrile illnesses. These findings collectively suggest that the NINJ1-HMGB1 axis may offer novel insights into the mechanisms underlying KD vasculitis, supporting further investigation into its potential clinical relevance. Full article

Background: Pulmonary fibrosis (PF) and post-infectious fibrotic lung disease are characterized by marked heterogeneity in radiologic patterns, physiologic impairment, and clinical presentation. Conventional analytic approaches often fail to capture non-linear and multimodal relationships between structural imaging findings and functional limitation. Integrating imaging-derived representations with clinical and functional data using artificial intelligence (AI) may provide a more comprehensive characterization of disease heterogeneity. Objectives: The objective of this study was to develop and evaluate a multimodal AI framework combining imaging-derived embeddings and structured clinical data to identify atypical clinical–radiologic profiles in patients with pulmonary fibrosis using unsupervised anomaly detection. Methods: A retrospective cohort of 41 patients with radiologically confirmed pulmonary fibrosis or post-infectious fibrotic lung disease was analyzed. Deep imaging embeddings were extracted from baseline thoracic CT examinations using a pretrained convolutional neural network and integrated with standardized clinical and functional variables. A multimodal variational autoencoder (VAE) was trained in an unsupervised manner to learn the distribution of typical patient profiles. Patient-specific anomaly scores were derived from reconstruction error plus latent regularization (β·KL divergence). Associations between anomaly scores, disease severity, and clinical markers were assessed using Spearman rank correlation. Results: Anomaly scores were right-skewed (median 26.91, IQR 22.87–32.11; range 19.75–46.18). Patients above the 85th percentile (anomaly score ≥ 33.85) comprised 7/41 (17.1%) of the cohort and occurred across all clinician-assigned severity categories (mild 3, moderate 1, severe 3). Anomaly scores overlapped substantially across severity groups, with similar medians (mild 26.47, moderate 28.55, severe 28.23). Correlations with conventional severity markers were weak and non-significant, including DLCO (% predicted; ρ = −0.25, p = 0.115) and FEV₁ (% predicted; ρ = −0.22, p = 0.165), a pattern consistent with anomaly scores reflecting multimodal deviation rather than severity alone, while acknowledging the exploratory nature of the analysis. Highly anomalous patients frequently exhibited discordant clinical–radiologic profiles, including preserved functional capacity despite marked imaging-derived deviation or disproportionate physiological impairment relative to imaging patterns. Conclusions: This proof-of-concept study demonstrates that multimodal VAE-based anomaly detection integrating imaging-derived embeddings with clinical data can quantify clinical–radiologic heterogeneity in pulmonary fibrosis beyond conventional severity stratification. Unsupervised anomaly detection provides a complementary framework for identifying atypical multimodal profiles and supporting individualized phenotyping and hypothesis generation in fibrotic lung disease. Given the modest cohort size, these findings should be interpreted as illustrative and hypothesis-generating rather than generalizable. Full article

Distant metastasis is the leading cause of breast cancer-related mortality, yet its prediction from primary tumor profiles remains challenging. Cytoskeletal remodeling and cell motility are central to metastatic dissemination, suggesting mechanobiological genes as biologically relevant biomarkers. Here, we evaluated the ability of supervised machine learning models to distinguish metastatic from non-metastatic breast cancer samples using expression profiles of 11 actin cytoskeleton-related genes from the TCGA cohort. Ensemble models, particularly Random Forest and XGBoost, demonstrated strong discriminative ability and consistently outperformed other approaches after employing SMOTE for class balancing in exploratory analyses. CFL1, ANXA2, and MYH9 consistently emerged as the most informative predictors, highlighting mechanobiological processes as key drivers of metastatic risk. Full article

Background: Anastomotic leaks (ALs) remain a critical complication after rectal cancer surgery. Emerging evidence suggests that local immune dysregulation may play a key role in anastomotic healing. We investigated the immune microenvironment of histologically normal, tumor-adjacent rectal mucosa—a tumor-conditioned field—as a potential substrate for AL predisposition. Methods: IMMUNOREACT 4 is a sub-analysis of the IMMUNOREACT project (clinicaltrials.gov NCT04915326 and NCT04915326), a multicenter translational study evaluating immune features of histologically normal, tumor-adjacent rectal mucosa of patients undergoing colorectal anastomosis. A prospective cohort (n = 121) was analyzed using flow cytometry, in addition to a retrospective cohort (n = 262) using immunohistochemistry. Immune markers of epithelial activation and lymphocyte subsets were compared between patients with and without postoperative ALs. Exploratory predictive models combining immune and clinical variables were developed and evaluated using discrimination, calibration and decision curve analyses. Results: At flow cytometry, the CK⁺HLAabc⁺ MFI (AUC 0.66, 95% CI 0.52–0.80), CD8⁺CD38⁺ cell rate (AUC 0.65, 95% CI 0.52–0.78) and CD3⁺CTLA4⁺ cell rate (AUC 0.65, 95% CI 0.51–0.80) showed moderate predictive potential for ALs. In immunohistochemistry, CD3⁺ (AUC 0.57, 95% CI 0.54–0.60), CD8⁺ (AUC 0.57, 95% CI 0.52–0.62), CD8β⁺ (AUC 0.59, 95% CI 0.53–0.65) and Tbet⁺ (AUC 0.60, 95% CI 0.56–0.64) showed some predictive ability for ALs. The model including CD8β⁺, the BMI, neutrophile/lymphocyte ratio and tumor location had an AUC of 0.67 (95% CI 0.62–0.72). Conclusions: Immune activation within histologically normal, tumor-adjacent rectal mucosa—characterized by epithelial HLA upregulation and cytotoxic or Th1 T cell infiltration—is associated with postoperative ALs. Although predictive accuracy is limited, these findings support the concept that a tumor-conditioned immune microenvironment may predispose patients to impaired anastomotic healing. Integration of mucosal immune profiling with clinical variables represents a promising exploratory approach that warrants further prospective validation. Full article

Background/Objectives: Delayed cerebral ischemia (DCI) is a major cause of poor outcome after aneurysmal subarachnoid hemorrhage (aSAH). Beyond large-vessel vasospasm, DCI reflects a systemic, multifactorial process involving inflammation, hematologic dysregulation, and organ dysfunction. Stroke-associated pneumonia (SAP), a frequent aSAH complication linked to stroke-induced immunodepression, may aggravate secondary ischemic injury. Unlike prior studies focusing on classical predictors alone, we included pneumonia and longitudinal respiratory parameters alongside inflammatory, hematologic, and renal markers. Using machine learning, this study aimed to identify predictors of DCI and functional outcome from routinely collected intensive care data. Methods: In this retrospective single-center study, 182 aSAH patients treated in a neurosurgical intensive care unit were included. Clinical data, SAP status, and longitudinal inflammatory, hematologic, renal, and respiratory parameters were extracted. DCI and functional outcome were assessed. Continuous variables were summarized as minimum, maximum, and mean values. Supervised machine learning models combining 12 feature selection methods and 12 classifiers were trained using five-fold cross-validation and evaluated by accuracy, F1-score, and AUC. Results: DCI occurred in 22% of patients, and SAP in 27%. The machine learning models achieved a mean accuracy of 59.7% (F1-score 58.8%, AUC 59.7%) for DCI prediction. No single dominant feature emerged; predictive patterns included leukocyte counts, CRP, erythrocyte indices, platelet variability, renal function, and oxygenation metrics. Functional outcome prediction performed moderately better (mean AUC 65.7%) and shared overlapping predictors. Conclusions: DCI reflects systemic instability in aSAH, with longitudinal inflammatory and respiratory variability outperforming static thresholds. Dynamic risk stratification may enable earlier detection of deterioration, supporting future time-series modeling and external validation. Full article

Background: Cardiac amyloidosis (CA) is mainly characterized by diastolic dysfunction, with gradually worsening functional capacity and poor prognosis. Left atrial (LA) strain by speckle tracking echocardiography (STE) is an index of diastolic function and heart failure (HF) symptoms. The aim of this study was to evaluate the relationship of LA strain with functional capacity in CA and the potential prognostic value of speckle tracking variables. Methods: In this single-center study, we prospectively enrolled consecutive outpatients with CA (n = 75). Clinical, echocardiographic evaluation, six-minute-walking-test (6MWT) and Kansas City Cardiomyopathy Questionnaire (KCCQ) were performed on the same day. The primary endpoint was the correlation between global peak atrial longitudinal strain (PALS) and NTproBNP, 6MWT score, and KCCQ. The secondary endpoint was a combination of all-cause or cardiovascular death and HF hospitalization. Results: Overall, 48 ATTR and 27 AL patients (74 ± 11 years, 84% male) were enrolled. Global PALS showed a significant direct correlation with N-terminal-pro-brain natriuretic peptide (NTproBNP, p = 0.3, p = 0.017) and 6MWT (p = 0.4, R² = 0.2, p = 0.004), but no significant correlation with KCCQ (p = −0.13, p = 0.3). GLS showed a significant direct correlation with NTproBNP (p = 0.3, p = 0.017) but not with 6MWT and/or KCCQ. Over a mean follow up of 12 ± 3 months, 42 patients reached the combined endpoint. With ROC curves, both global PALS < 13.5% and GLS > −12% provided a good prediction of the combined endpoint (AUC = 0.72 [0.6–0.82] and 0.73 [0.63–0.83], respectively, p < 0.0001), higher than NTproBNP and other echocardiographic parameters. Conclusions: Global PALS is associated with congestion and functional capacity in CA, suggesting its role as a more objective marker of disease severity in CA. Speckle tracking parameters may be used to enhance prognostic stratification in CA. Full article

Large language models (LLMs) are now routine writing tools across various domains, intensifying questions about when text should be treated as human-authored, artificial intelligence (AI)-generated, or collaboratively produced. This rapid review aims to identify cue families reported in empirical studies as distinguishing AI from human-authored text and to assess how stable these cues are across genres/tasks, text lengths, and revision conditions. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines, we searched four online databases for peer-reviewed empirical articles (1 January 2022–1 January 2026). After deduplication and screening, 40 studies were included. Evidence converged on five cue families: surface, discourse/pragmatic, epistemic/content, predictability/probabilistic, and provenance. Surface cues dominated the literature and were the most consistently operationalized. Discourse/pragmatic cues followed, particularly in discipline-bound academic genres where stance and metadiscourse differentiated AI from human writing. Predictability/probabilistic cues were central in detector-focused studies, while epistemic/content cues emerged primarily in tasks where grounding and authenticity were salient. Provenance cues were concentrated in watermarking research. Across studies, cue stability was consistently conditional rather than universal. Specifically, surface and discourse cues often remained discriminative within constrained genres, but shifted with register and discipline; probabilistic cues were powerful yet fragile under paraphrasing, post-editing, and evasion; and provenance signals required robustness to editing, mixing, and span localization. Overall, the literature indicates that AI–human distinction emerges from layered and context-dependent cue profiles rather than from any single reliable marker. High-stakes decisions, therefore, require condition-aware interpretation, triangulation across multiple cue families, and human oversight rather than automated classification in isolation. Full article

Cystic fibrosis (CF) is characterized by neutrophil-driven airway inflammation and acute respiratory events (AREs) that contribute to progressive lung damage. AREs are clinically heterogeneous and often occur without measurable changes in lung function. This study aimed to evaluate the utility of molecular airway inflammatory markers for detecting AREs in school-age children with CF. We performed a secondary analysis of a prospective observational study of children with CF (ages 6.7–16.8 years) followed for two years. Sputum samples were collected from 50 participants during stable visits and AREs. Concentrations of 14 inflammatory cytokines were measured using ELISA and multiplex assays. Associations with lung function (ppFEV₁ and lung clearance index [LCI]) and time to next ARE were assessed. A total of 179 sputum samples were analyzed, including 64 collected during AREs. Calprotectin, interleukin-8 (IL-8), and IL-1β were increased during AREs compared with stable visits, although concentrations frequently remained within ranges observed at stable visits. Other cytokines, including GM-CSF, IL-17A, IL-1α, TNF-α, and SPLUNC-1, were predictive of shorter time to subsequent AREs. No biomarker correlated with lung function measures. These findings indicate that airway inflammatory cytokine changes are associated with clinically diagnosed AREs but not with pulmonary function, supporting their potential role as complementary biomarkers in CF care. Full article

Alzheimer’s disease (AD) is the most common cause of dementia and is characterized by progressive cognitive decline, β-amyloid accumulation, tau pathology, oxidative stress, and neuroinflammation. Increasing evidence suggests that metabolic dysregulation may contribute to AD pathogenesis. Glucagon-like peptide-2 (GLP-2), an intestinal peptide hormone, has demonstrated neuroprotective effects in preclinical models, potentially through anti-inflammatory and anti-apoptotic mechanisms. However, its role in human neurodegenerative disorders remains insufficiently understood. This study aimed to compare plasma GLP-2 concentrations between individuals with AD and cognitively healthy controls and to examine associations between GLP-2 levels, cognitive impairment severity, and metabolic parameters. Sixty-one patients with clinically diagnosed AD and twenty-three cognitively unimpaired controls were recruited. Plasma total GLP-2 concentrations were assessed at baseline in all participants and additionally at 6 and 12 months in a subgroup of 34 AD patients. Cognitive function was evaluated using the Mini-Mental State Examination (MMSE) and the Clinical Dementia Rating (CDR) scale. Group comparisons, subgroup analyses based on AD severity, repeated-measures analyses, Spearman correlations, and multivariable linear regression models (including age and clinical group) were performed. Plasma GLP-2 concentrations were significantly higher in AD patients than in controls, with a moderate effect size (Cohen’s d ≈ 0.60). In severity-based subgroup analyses, both the mild and moderate-to-severe AD groups showed significantly higher GLP-2 levels than controls. Longitudinal analyses in AD patients (n = 34) showed no significant changes in GLP-2 concentrations over 12 months. Cognitive performance declined over time, with a significant reduction in MMSE from baseline to 6 months, whereas GLP-2 levels were not correlated with MMSE or CDR at any time point. GLP-2 levels correlated positively with body mass index (BMI), body weight, insulin, and HOMA-IR. In multivariable regression analysis, neither age nor clinical group independently predicted GLP-2 concentrations (both p > 0.05). Plasma GLP-2 concentrations were higher in patients with AD than in cognitively healthy controls; however, GLP-2 levels were not associated with cognitive performance or its progression over 12 months. GLP-2 was positively related to markers of adiposity and insulin resistance, suggesting stronger links to metabolic status than to cognitive severity. Further studies are needed to clarify whether GLP-2 alterations in AD reflect compensatory mechanisms, metabolic factors, or disease-related pathophysiology. Full article

Background: Early mortality risk stratification is essential in acute pulmonary embolism (PE). Integrating clinical variables with biomarkers may enhance prognostic accuracy beyond established tools. Methods: In a retrospective cohort of 322 patients with confirmed acute PE, we evaluated syncope, right-ventricular (RV) dysfunction, systolic blood pressure (SBP) and biochemical markers as candidate predictors of in-hospital mortality. A penalized logistic regression model using LASSO (least absolute shrinkage and selection operator) was developed and internally validated with five-fold cross-validation and 200 bootstrap repetitions. Discrimination, calibration and clinical utility were assessed using the area under the receiver operating characteristic curve (AUC), Brier score and decision-curve analysis (DCA). Results: In-hospital mortality was 5.6% (n = 18). LASSO retained four predictors: syncope, RV dysfunction, lower SBP and higher troponin levels. The optimism-corrected AUC was 0.70 (95% CI 0.63–0.77), with strong calibration (Brier score 0.066). DCA showed that the model provided greater net benefit than treat-all, treat-none, and sPESI strategies across threshold probabilities of approximately 7–25%, supporting its potential value for early triage. NT-proBNP, D-dimer and lactate did not add incremental predictive information after penalization. Conclusions: A simple, interpretable model integrating clinical parameters and troponin demonstrates good predictive performance and favorable clinical utility for early mortality risk estimation in acute PE. External validation is required before broader implementation. Full article

Background/Objectives: Background parenchymal enhancement (BPE) on breast magnetic resonance imaging reflects hormonal and vascular activity of fibroglandular tissue and is studied as a prognostic marker for breast cancer. This paper serves as a review that evaluates quantitative methods for BPE measurements for predicting treatment outcomes. Methods: PubMed was searched for papers on evaluating BPE with outcomes to compare, such as pathologic complete response, recurrence-free survival, disease-free survival, and overall survival, from 2015 to 2025. In total, eleven papers using quantitative methods to measure BPE were selected. Results: Quantitative results showed that BPE reduction during neoadjuvant chemotherapy and high pre-treatment/baseline BPE are linked to improved treatment response and reduced risk of recurrence. Conclusions: Quantitative assessment methods yield objective and reproducible prognostic information. Incorporating quantitative BPE measurements alongside tumor-focused imaging features may further improve predictive accuracy in clinical settings. Full article

This study investigates the distinct mechanisms of human versus Large Language Model (LLM) creativity. Employing a two-stage experimental design, we systematically compared Human-Only, LLM-Only, and LLM-Assisted performance across propositional and creative writing tasks. Results revealed a critical asymmetry contingent upon the research context: human authors exhibited higher originality in high-demand creative tasks, whereas LLMs governed execution quality, maintaining superior effectiveness across different tasks and cohorts. This pattern is characterized by four exploratory writing creativity profiles: Ideal, Safe, Moderate, and Plain. The distribution of human and LLM writings across these profiles was strikingly different. Hierarchical Moderated Regression analysis uncovered divergent linguistic pathways: human originality is predicted by markers of subjective cognitive investment, while LLM effectiveness is mechanistically driven by optimized structural coherence. Furthermore, the study identified a “Collaboration Trap” during collaboration with a suboptimal LLM. This partnership failed to improve human performance relative to LLM-Only benchmarks and induced cognitive anchoring, leading humans to mimic AI complexity without quality gains. These insights offer critical implications for preserving human agency and avoiding homogenization in future human–AI collaborative writing pedagogies. Full article

Background/Objectives: Virus-like particles (VLPs) are effective vaccine platforms but are susceptible to degradation, which compromises stability and immunogenicity. A key challenge is the lack of sensitive early indicators of instability. This study aimed to systematically evaluate the stability of an aluminum-free recombinant hepatitis E virus VLP vaccine under various stresses and identify predictive markers of instability. Methods: The VLP vaccine was subjected to thermal stress (4 °C, 25 °C, 37 °C, 56 °C for up to 28 d), repeated freeze–thaw cycles (up to 30 cycles), and mechanical agitation (orbital shaking at 100 and 300 rpm for up to 12 d). Stability was assessed using a multi-parameter panel monitoring critical quality attributes: conformational and colloidal stability, formation of high-molecular-weight species, mean particle size, polydispersity index, charge heterogeneity, and antigen content. Results: Changes in charge heterogeneity were the earliest indicator of instability, detectable within 3 days at 25 °C, 8 h at 37 °C, and 4 h at 56 °C, preceding losses in structural integrity or antigen-binding function. The VLPs remained stable at 25 °C for 28 d. Freeze–thaw cycles induced a basic shift in charge variants without compromising structure or function, while high-intensity shaking (300 rpm) caused aggregation after 3–6 d. The effects of common excipients were also characterized. Conclusions: Charge-variant analysis serves as a sensitive and predictive marker for VLP vaccine instability. The study delineates the distinct impacts of different stress factors and provides critical data for optimizing formulation design and storage strategies to enhance VLP vaccine stability. Full article

Cold atmospheric plasma (CAP) has emerged as a promising anticancer approach because of its ability to selectively eliminate malignant cells. Among the proposed mechanisms of this selectivity, the Bauer theory emphasizes the synergistic action of plasma-derived hydrogen peroxide (H₂O₂) and nitrite (NO₂⁻), leading to the transient generation of primary singlet oxygen (¹O₂). This early event inactivates membrane-bound catalase, allowing tumor cell-derived H₂O₂ and peroxynitrite to initiate a self-amplifying cycle that produces secondary ¹O₂, as a hallmark of CAP selectivity. To test this hypothesis, in this work, we monitored extracellular dissolved oxygen (DO) dynamics in HT-29 colorectal cancer cells treated with an argon plasma jet using time-resolved phosphorescence lifetime spectroscopy. Temporal variations in DO likely reflect the cumulative effect of rapid ¹O₂ production and its reactions with cells. A delayed surge in extracellular ¹O₂ was observed specifically in dying cancer cells within the 10–20 min window predicted by the model. Intracellular ROS imaging confirmed a strong correlation between intracellular ROS, extracellular ¹O₂ dynamics, and viability loss. Together, these results provide mechanistic validation of Bauer’s redox model and suggest that early oxygen dynamics after CAP exposure can serve as predictive markers for treatment efficacy in plasma or photodynamic therapies. Full article