Abstract Background This study aims to examine the independent relationships between individual components of metabolic syndrome (MetS) and two key clinical outcomes in patients with Crohn’s disease (CD): disease activity, as quantified by the Crohn’s Disease Activity Index (CDAI), and the occurrence of complications. Methods This retrospective cross-sectional study included 376 adults with newly diagnosed Crohn’s disease. Multiple linear regression was used to examine associations between metabolic parameters and CDAI scores, while multivariate logistic regression assessed links to complications. Analyses were also based on clinical CDAI cut-offs. Predictive nomograms were developed and internally validated via bootstrap resampling. Results Multiple linear regression indicated that higher CDAI scores were independently associated with lower BMI (B = −5.866, P < 0.001), lower HDL-C levels (B = −81.770, P < 0.001), higher triglycerides (B = 15.618, P = 0.001), and lower ESR (B = −0.375, P = 0.03). Multivariate logistic regression established low HDL-C (OR = 0.042, P < 0.001), low BMI (OR = 0.915, P = 0.034), and high triglycerides (OR = 1.792, P = 0.007) as significant independent risk factors for complications. The developed nomograms demonstrated strong predictive performance, with an adjusted R 2 of 0.207 for the CDAI model and an AUC of 0.765 for the complication model. For both predictive tasks, the model incorporating separate TG and HDL-C measurements significantly outperformed the TG/HDL-C ratio model. Conclusion Metabolic disturbances demonstrate a significant association with increased disease severity and a higher risk of complication development in Crohn’s disease. Core tip Dual-outcome study reveals HDL-C and TG differentially link to CD inflammation and complications, pointing to distinct mechanisms. Low HDL-C is the strongest independent predictor for CD complications, underscoring its protective role beyond cholesterol transport. Individual TG and HDL-C metrics outperform their ratio in prediction, challenging its use and suggesting independent pathways in CD. Low BMI independently associates with both adverse outcomes, refining the “obesity paradox” and highlighting malnutrition’s key role. A practical, validated nomogram (AUC=0.765) integrates HDL-C, TG, and BMI to stratify complication risk, aiding clinical decision-making.

Molecular heterogeneity of gastric cancer leads to differences in treatment responses and clinical outcomes. Molecular subtype classification, based on gene expression patterns, provides insights into disease mechanisms and potential therapeutic targets. Electrochemical sensors have significant application value in the field of medical testing. They can achieve rapid and sensitive detection of multiple biomarkers by detecting changes in electrochemical signals in biological samples, providing support for the early diagnosis, treatment monitoring and prognosis assessment of diseases. In recent years, electrochemical sensor technology has made significant progress in detecting cancer-related biomarkers, providing new means for the molecular subtype classification and personalized treatment of gastric cancer. This study aims to identify the molecular subtypes of gastric cancer and their specific marker genes through comprehensive bioinformatics analysis and electrochemical sensor technology, and further explore the application potential of electrochemical sensors in the classification of molecular subtypes of gastric cancer. The study obtained gastric cancer gene expression data from the gene expression database, identified molecular subtypes by consensus clustering methods, and verified subtype separation through multiple dimension reduction techniques. Biomarkers related to tumor necrosis factor response and actin tissue were detected. Combined with bioinformatics analysis, molecular subsub-specific marker genes were identified. The functional characteristics of each subtype were revealed through methods such as differential expression analysis and functional enrichment analysis. Through consensus clustering analysis, three different molecular subtypes of gastric cancer were identified. Among them, Subtype1 mainly contains adenocarcinoma samples, Subtype2 mainly contains normal tissue samples, and Subtype3 is also mainly composed of adenocarcinoma samples. The combination of biomarker expression data detected by electrochemical sensors and gene expression data further verified the accuracy of molecular subtype classification. Differential expression analysis revealed that there were 1100 genes differentially expressed between Subtype1 and Subtype2, and 200 core genes shared in the comparison were identified. Functional enrichment analysis revealed subtype-specific tumor necrosis factor responses, activation of actin filament tissue and metabolic pathways. Electrochemical sensors have unique advantages in detecting biomarkers related to gastric cancer. Their high sensitivity and specificity enable them to detect low-abundance biomarkers, providing the possibility for the fine classification of molecular subtypes.

BACKGROUND:Chemotherapy resistance remains a major challenge in pancreatic cancer treatment, with gemcitabine-based therapy being a primary approach. However, the mechanisms underlying gemcitabine resistance in pancreatic cancer cells are not yet fully understood. This study aimed to investigate the role ofHBE1in conferring resistance to gemcitabine and explore its effects on reactive oxygen species (ROS) production, mitochondrial apoptosis, and endoplasmic reticulum (ER) stress. MATERIALS AND METHODS:A gemcitabine-resistant pancreatic cancer cell line was systematically developed through gradual, six-month exposure to increasing concentrations of gemcitabine. The role ofHBE1was examined by assessing its impact on ROS levels, mitochondrial-mediated apoptosis, and ER stress pathways, includingIRE1α phosphorylationandGRP78 expression. RESULTS:Overexpression ofHBE1was found to suppress ROS generation and inhibit mitochondrial-dependent apoptosis. Additionally, it attenuated ER stress by reducingIRE1α phosphorylationwhile upregulatingGRP78. CONCLUSION:These findings demonstrate a strong correlation between elevatedHBE1expression and gemcitabine resistance in pancreatic cancer cells, suggesting its potential as a therapeutic target to overcome chemoresistance.

Insulinoma is a rare functional pancreatic neuroendocrine tumor characterized by excessive insulin secretion, which causes hypoglycemia. Recent advances in endoscopic technology have provided essential tools for the diagnosis and treatment of insulinoma. Early diagnosis and timely intervention, particularly endoscopy-guided intervention in selected patients, may improve clinical outcomes and quality of life. Although existing national and international guidelines mention insulinoma, standardized criteria for its endoscopic diagnosis and treatment remain lacking. Therefore, under the leadership of the Digestive Endoscopy Branch of the Chinese Medical Association, experts from related fields were invited to develop the Expert Consensus on Endoscopic Diagnosis and Treatment for Insulinoma (2026). The development process included a comprehensive review of recent national and international evidence-based literature, with the quality of evidence and strength of recommendations assessed according to the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system. This consensus emphasizes endoscopy-guided diagnostic and therapeutic approaches. It aims to provide practical recommendations and operational guidance for standardized endoscopic diagnosis and treatment of insulinoma, thereby advancing the quality of endoscopic care for this condition in China.

Background GPR84 is involved in inflammation and metabolism, but its function in CRC pathogenesis remains unclear. This study investigated the functional interaction between GPR84 and the adaptor protein DOK3 in CRC progression. Methods Comprehensive analyses were performed by integrating multi-omics data from TCGA, GTEx, and GEO cohorts, along with single-cell and spatial transcriptomics datasets, with functional enrichment conducted to explore relevant biological pathways. CRC tissues were examined via immunohistochemistry. In vitro functional assays (qRT-PCR, Western blot, Co-IP, Transwell, CCK-8) were conducted in HCT116 and SW480 cells following GPR84 activation with agonist 6-OAU or DOK3 silencing. Results GPR84 was significantly upregulated in CRC tissues and showed moderate diagnostic value (AUC = 0.738). High GPR84 expression correlated with specific molecular subtypes, enhanced immune signatures, and elevated tertiary lymphoid structure scores. Single-cell analysis revealed predominant GPR84 expression in myeloid cells, strongly correlating with DOK3 expression (R = 0.69, p < 2.2 & times; 10(-16)). 6-OAU treatment suppressed CRC cell migration, invasion, and proliferation, effects reversed by DOK3 knockdown. Co-IP confirmed direct GPR84-DOK3 interaction. Conclusion The GPR84-DOK3 axis represents a novel regulatory pathway in CRC, where GPR84 activation suppresses malignant behaviors through DOK3, highlighting its potential as a therapeutic target and biomarker.

Objectives This study aimed to develop interpretable deep learning (DL) and radiomics models using endoscopic ultrasound (EUS) images to differentiate insulinomas from nonfunctional pancreatic neuroendocrine tumors (NF-PNETs). Methods The retrospective analysis comprised 115 patients, including 61 with insulinomas and 54 with NF-PNETs, all confirmed through pathological examination. The patient cohort was divided into training and test groups. From standardized EUS images, a total of 512 DL features and 107 radiomics features were extracted. LASSO regression was employed to identify non-zero coefficient features from both the DL and radiomics datasets. Subsequently, four machine learning algorithms were utilized to construct predictive models. The optimal DL and radiomics models were then integrated into a nomogram for enhanced predictive capability. Gradient-weighted Class Activation Mapping (Grad-CAM) and Shapley Additive Explanations (SHAP) provided model interpretability. Results The ExtraTrees DL and radiomics models demonstrated exceptional performance. The integrated nomogram yielded AUC values of 0.978 for the training group and 0.842 for the test group. Calibration curves and decision curve analysis corroborated the high accuracy and clinical utility of the models. Grad-CAM identified tumor margins and heterogeneity as significant features in the DL model, whereas SHAP analysis highlighted texture patterns in the radiomics data.The nomogram effectively facilitated visual simplification of risk stratification. Conclusions The interpretable DL-radiomics nomogram exhibited significant potential in differentiating insulinomas from NF-PNETs using EUS.This methodology improves diagnostic accuracy and informs clinical decision-making.

Concomitant with the escalating global burden of metabolic disorders, acute pancreatitis with hypertriglyceridemia (AP-HTG) has become highly prevalent. However, the initial triglyceride elevation does not strictly correlate with adverse clinical outcomes. Therefore, we aimed to identify novel dynamic subphenotypes reflecting the intrinsic pathophysiological state and develop an early prediction model. Clinicopathological data were collected from MIMIC-IV (n = 344, development), eICU-CRD (n = 173, external validation), and an independent local cohort (n = 319, pragmatic validation). Group-based multi-trajectory modeling (GBMTM) of 7-day white blood cell and calcium trajectories was utilized to identify novel subphenotypes. After evaluating multiple machine-learning algorithms, an optimal early prediction model was established using Boruta, LASSO, and logistic regression based on admission-day indicators. Four novel dynamic subphenotypes were identified, with the Severe-Hyperinflammatory (C3) and Lipotoxic-Hypocalcemic (C4) subphenotypes carrying substantially increased independent risks for persistent organ failure and multiple organ dysfunction syndrome. Furthermore, an early logistic regression prediction model comprising white blood cells, calcium, lymphocytes, albumin, platelets, and hematocrit was developed. This model achieved high macro-average AUCs in the development (0.924) and external validation (0.933) cohorts. Pragmatic validation in the local cohort further confirmed its utility in effectively identifying high-risk subphenotypes. To enhance interpretability and clinical utility, SHAP analysis was applied, and the model was deployed as a free web-based calculator and a nomogram specifically designed for the C4 subphenotype. This study identified four novel dynamic AP-HTG subphenotypes and developed a multicenter-validated early prediction model using admission-day indicators. It facilitates early risk stratification and individualized clinical decision-making.

INTRODUCTION:Sound speed correction endoscopic ultrasound (SSC-EUS) is a novel imaging technique with limited previous validation. The aim of this study was to evaluate the diagnostic efficacy of SSC-EUS for solid pancreatic lesions (SPL) and compare it with B-mode endoscopic ultrasound (B-EUS), elastography endoscopic ultrasound (EG-EUS), and contrast-enhanced endoscopic ultrasound (CE-EUS). METHODS:A prospective, single-blind, randomized trial included 240 patients with computed tomography/magnetic resonance imaging-confirmed SPL (solid portion >80% of lesion volume). Participants were equally divided into 4 groups (B-EUS, EG-EUS, CE-EUS, and SSC-EUS). Diagnostic thresholds were determined through receiver operating characteristic curves. Subgroup analyses assessed the impact of lesion location (head/body/tail), tumor size (≤3 vs >3 cm), and cancer stage (I/II vs III/IV). Statistical analysis used SPSS 23.0 and GraphPad Prism 8. RESULTS:Among 240 patients, 138 (57.5%) had malignant lesions. SSC-EUS achieved optimal diagnostic performance at a cutoff sound speed of 1,563 m/s (area under the receiver operating characteristic curve = 0.822, sensitivity = 82.8%, specificity = 78.9%, accuracy = 81.7%). CE-EUS demonstrated the highest overall efficacy (sensitivity = 90.3%, specificity = 82.8%, accuracy = 86.7%), followed by SSC-EUS and EG-EUS, both outperforming B-EUS (accuracy = 70.0%). Subgroup analysis revealed superior sensitivity for pancreatic body lesions (SSC-EUS: 87.5%; CE-EUS: 90.0%), tumors >3 cm (SSC-EUS: 84.2%; CE-EUS: 90.0%), and stage III/IV cancers (SSC-EUS: 81.8%; CE-EUS: 90.9%). EG-EUS strain ratio (cutoff = 4.44) showed limited accuracy (61.7%), whereas elastic strain value A (cutoff = 0.065%) exhibited moderate utility (accuracy = 75.0%). DISCUSSION:CE-EUS remains the most effective imaging modality for SPL diagnosis. SSC-EUS demonstrates comparable accuracy with EG-EUS and is particularly advantageous for larger tumors (>3 cm) and advanced-stage malignancies. EG-EUS strain ratio lacks clinical robustness, whereas elastic strain value A warrants further validation. Tailoring imaging method selection to lesion characteristics (location, size, and stage) may optimize diagnostic outcomes.

This study aims to develop and validate an interpretable deep learning (DL) model and a nomogram based on endoscopic ultrasound (EUS) images for the prediction of pathological grading in pancreatic neuroendocrine tumors (PNETs). This multicenter retrospective study included 108 patients with PNETs, who were divided into train (n = 81, internal center) and test cohorts (n = 27, external centers). Univariate and multivariate logistic regression were used for screening demographic characteristics and EUS semantic features. Deep transfer learning was employed using a pre-trained ResNet18 model to extract features from EUS images. Feature selection was conducted using the least absolute shrinkage and selection operator (LASSO), and various machine learning algorithms were utilized to construct DL models. The optimal model was then integrated with clinical features to develop a nomogram. The performance of the model was assessed using the area under the curve (AUC), calibration curves, decision curve analysis (DCA), and clinical impact curves (CIC). The nomogram, which integrates the optimal DL model (Naive Bayes) with clinical features, achieved AUC values of 0.928 (95

BACKGROUND:Acute pancreatitis (AP) is a disease with high morbidity and mortality. Neutrophils are highly correlated with the occurrence and severity of the inflammatory response in AP. However, the mechanism by which neutrophils act on AP is currently unclear. We aimed to target genes coexpressed with neutrophils to identify new possibilities for diagnosing and treating AP. METHODS:We analyzed the differences in immune infiltration in AP and the differential expression of neutrophil-coexpressed genes. Then, models were built through various machine learning methods, and the hub neutrophil coexpressed genes were identified. Receiver operating characteristic (ROC) curves were used for validation, and the relationships between the hub neutrophil coexpressed genes and immune infiltration were explored. GSEA and GSVA were used to investigate the biological functions of genes. Finally, we explored the miRNAs and lncRNAs associated with the hub neutrophil coexpressed genes. RESULTS:Neutrophil infiltration varied significantly in samples from patients with AP. RGS2, AHNAK, and OGT were identified as hub neutrophil coexpressed genes by taking intersections of genes involved in the machine learning approach. The hub neutrophil coexpressed genes are closely associated with a variety of immune cells. GSVA revealed that the hub neutrophil coexpressed genes are differentially expressed in multiple immune- and inflammation-related pathways. These findings suggest that these genes may influence AP progression through these pathways. CONCLUSIONS:In this study, genes coexpressed with neutrophils in AP were identified, and their biological functions were investigated. These findings may provide more effective therapeutic strategies for the prediction, prevention, and personalized treatment of patients with AP.

BackgroundThe impact of metabolic syndrome (MetS) and its components on Crohn’s disease (CD) remains unclear. This study investigated how individual MetS factors and cumulative metabolic burden affect CD activity and outcomes.MethodsThis retrospective study was conducted at a tertiary care hospital and encompassed a cohort of 376 hospitalized patients diagnosed with CD from 2015 to 2025. Linear, logistic, and Poisson regression models assessed correlations between MetS elements and clinical indicators, evaluating how their presence and number affected CD.ResultsThirty-six patients (9.6%) had MetS. Compared to non-MetS patients, those with MetS exhibited a significantly higher Simple Endoscopic Score for CD (10 vs. 7; p < 0.001), Crohn’s Disease Activity Index (294.3 vs. 256.6; p < 0.001), risks of complications (OR = 8.65, 95% CI: 2.01–37.26, p = 0.004) and surgery or invasive procedures (OR = 2.64, 95% CI: 1.27–5.45, p = 0.009). Low high-density lipoprotein cholesterol conferred a higher risk of adverse outcomes. The cumulative number of MetS elements exhibited an incremental effect, with increasing numbers correlating to progressively higher disease severity and a risk of poor outcomes.ConclusionThe concurrent presence of multiple MetS elements can synergistically worsen the clinical course of CD. Management of these components is crucial for the long-term prognosis of CD.

ABSTRACT Purpose Rapid progression in late‐stage is a characteristic of pancreatic cancer (PC), leading to mortality. The critical role of lncRNA A2M‐AS1 (long non‐coding RNA alpha‐2‐macroglobulin antisense RNA 1) is involved in cancer progression, but the upstream regulator of A2M‐AS1 in the PC progression phenotype remains elusive. Methods We conducted an integrated analysis using bioinformatics, in vitro experiments, and in vivo studies. Human PC tissues were analyzed for A2M‐AS1 and p53 expressions. The PANC‐1 and BxPC‐3 cell lines were used for functional assays, including cell proliferation, apoptosis, migration, and invasion assays. The role of p53 in regulating A2M‐AS1 was investigated through overexpression and knockdown studies, along with using a MAPK pathway inhibitor. Results We found that A2M‐AS1 is downregulated in PC tissues and that its high expression correlates with a better prognosis. p53 was identified as a negative regulator of A2M‐AS1, with its knockdown leading to increased A2M‐AS1 expression and decreased PC cell invasiveness. Mechanistically, p53 was shown to bind to the A2M‐AS1 promoter, modulating its transcriptional activity. The MAPK pathway was revealed as a downstream effector of the p53‐A2M‐AS1 axis, with its inhibition reversing the effects on PC cell behavior. Conclusion High A2M‐AS1 expression is associated with a better PC prognosis. A2M‐AS1 overexpression subdues PC cell development. Our study unveils a new mechanism by which p53 decreases A2M‐AS1 expression.

Resveratrol alleviates liver fibrosis in mice by upregulating IL-10 to reprogram the macrophage phenotype; however, the mechanism remains to be elucidated. Building on our previous work, in this study, we aimed to determine the role of the TLR2/MyD88/ERK and NF-κB/NLRP3 inflammasome pathways in mediating the effects of resveratrol on liver fibrosis and macrophage polarization. We investigated the expression of cytokines in these inflammasome pathways in a mouse model of liver fibrosis and resveratrol-treated macrophages. The results showed that expression of TLR2, MyD88, ERK, and NF-κB1 in liver tissues was increased in the fourth week after treatment with resveratrol but decreased in the fifth week. Similar results were obtained for the NF-κB/NLRP3 inflammasome pathway. The results also showed that cytokines in both the TLR2/MyD88/ERK and NF-κB/NLRP3 inflammasome pathways in macrophages were elevated after 24 h and reduced after 36 h of treatment with resveratrol. Immunofluorescence and nucleocytoplasmic separation assays showed that resveratrol inhibited the translocation of NF-κB from the cytoplasm to the nucleus in macrophages, and increased NF-κB1 was associated with inhibition of the TLR2/MyD88/ERK pathway. Silencing NF-κB1 increased the expression of TLR2, MyD88, and ERK. In conclusion, the TLR2/MyD88/ERK and NF-κB/NLRP3 inflammasome pathways are involved in the effect of resveratrol on macrophage polarization and the subsequent modulation of liver fibrosis. NF-κB1 acts as the common cytokine that coordinates the crosstalk between the two pathways. Our findings highlight the potential of the members of these pathways as therapeutic targets toward the treatment of liver fibrosis.

To establish and validate a model based on CT imaging during follow-ups for predicting the disease progression in ileal stricturing Crohn’s disease (CD). Between January 2014 and February 2024, a retrospective review was conducted on 71 patients (training, n = 49; test, n = 22) who were initially diagnosed with ileal stricturing CD. Disease progression referred to the development of penetrating diseases, the requirement for CD-related hospitalization or surgery during follow-up. Radiomics features were extracted from visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) on baseline and follow-up CT scans, respectively. Integrating clinical characteristics and body composition features, a novel CT-based delta-radiomics nomogram was established according to multivariate Cox stepwise regression analysis. Receiver operating characteristic (ROC) analysis was performed to assess diagnostic performance. The delta-VAT radiomics model (RM) exhibited satisfactory performance in training cohort (the area under the ROC curve [AUC] = 0.792, 95

OBJECTIVE:To investigate the role of lncRNA Snhg6 in liver fibrosis, delivered by resveratrol-stimulated macrophage exosomes. METHODS:Resveratrol-stimulated and unstimulated exosomes were generated from RAW 264.7 cells, confirmed by electron microscopy, nanoparticle analysis, and Western blotting. JS1 cells were used as an HSC model, activated with TGF-β1 and treated with exosomes. Exosome uptake was observed via confocal microscopy, and acta2 expression was measured with immunofluorescence. RNA sequencing and RT-qPCR were used to analyze exosomal lncRNA profiles. KEGG GSEA enrichment was conducted on differentially expressed genes, and nf-κb expression was detected in HSCs using WB. Serum from liver fibrosis patients was analyzed for SNHG6 levels. RESULTS:Resveratrol-stimulated exosomes inhibited TGF-β1-induced HSC activation, with 132 differentially expressed lncRNAs, including upregulated Snhg6. NF-κB signaling was downregulated. Silencing Snhg6 weakened this inhibitory effect. CONCLUSION:Resveratrol-stimulated macrophage exosomes may inhibit liver fibrosis by delivering lncRNA Snhg6, which suppresses the NF-κB pathway.

To retrospectively develop and validate an interpretable deep learning model and nomogram utilizing endoscopic ultrasound (EUS) images to predict pancreatic neuroendocrine tumors (PNETs). Following confirmation via pathological examination, a retrospective analysis was performed on a cohort of 266 patients, comprising 115 individuals diagnosed with PNETs and 151 with pancreatic cancer. These patients were randomly assigned to the training or test group in a 7:3 ratio. The least absolute shrinkage and selection operator algorithm was employed to reduce the dimensionality of deep learning (DL) features extracted from pre-standardized EUS images. The retained nonzero coefficient features were subsequently applied to develop predictive eight DL models based on distinct machine learning algorithms. The optimal DL model was identified and used to establish a clinical signature, which subsequently informed the construction and evaluation of a nomogram. Gradient-weighted Class Activation Mapping (Grad-CAM) and Shapley Additive Explanations (SHAP) were implemented to interpret and visualize the model outputs. A total of 2048 DL features were initially extracted, from which only 27 features with coefficients greater than zero were retained. The support vector machine (SVM) DL model demonstrated exceptional performance, achieving area under the curve (AUC) values of 0.948 and 0.795 in the training and test groups, respectively. Additionally, a nomogram was developed, incorporating both DL and clinical signatures, and was visually represented for practical application. Finally, the calibration curves, decision curve analysis (DCA) plots, and clinical impact curves (CIC) exhibited by the DL model and nomogram indicated high accuracy. The application of Grad-CAM and SHAP enhanced the interpretability of these models. These methodologies contributed substantial net benefits to clinical decision-making processes. A novel interpretable DL model and nomogram were developed and validated using EUS images, cooperating with machine learning algorithms. This approach demonstrates significant potential for enhancing the clinical applicability of EUS in predicting PNETs from pancreatic cancer, thereby offering valuable insights for future research and implementation.

The objective is to develop and validate intratumoral and peritumoral ultrasomics models utilizing endoscopic ultrasonography (EUS) to predict pathological grading in pancreatic neuroendocrine tumors (PNETs). Eighty-one patients, including 51 with grade 1 PNETs and 30 with grade 2/3 PNETs, were included in this retrospective study after confirmation through pathological examination. The patients were randomly allocated to the training or test group in a 6:4 ratio. Univariate and multivariate logistic regression were used for screening clinical and ultrasonic characteristics. Ultrasomics is ultrasound-based radiomics. Ultrasomics features were extracted from both the intratumoral and peritumoral regions of conventional EUS images. Subsequently, the dimensionality of these radiomics features was reduced using the least absolute shrinkage and selection operator (LASSO) algorithm. A machine learning algorithm, namely multilayer perception (MLP), was employed to construct prediction models using only the nonzero coefficient features and retained clinical features, respectively. One hundred seven ultrasomics features based on EUS were extracted, and only features with nonzero coefficients were ultimately retained. Among all the models, the combined ultrasomics model achieved the greatest performance, with an AUC of 0.858 (95

ObjectivesThis study aimed to develop and validate intratumoral, peritumoral, and combined radiomic models based on endoscopic ultrasonography (EUS) for retrospectively differentiating pancreatic neuroendocrine tumors (PNETs) from pancreatic cancer.MethodsA total of 257 patients, including 151 with pancreatic cancer and 106 with PNETs, were retroactively enrolled after confirmation through pathological examination. These patients were randomized to either the training or test cohort in a ratio of 7:3. Radiomic features were extracted from the intratumoral and peritumoral regions from conventional EUS images. Following this, the radiomic features underwent dimensionality reduction through the utilization of the least absolute shrinkage and selection operator (LASSO) algorithm. Six machine learning algorithms were utilized to train prediction models employing features with nonzero coefficients. The optimum intratumoral radiomic model was identified and subsequently employed for further analysis. Furthermore, a combined radiomic model integrating both intratumoral and peritumoral radiomic features was established and assessed based on the same machine learning algorithm. Finally, a nomogram was constructed, integrating clinical signature and combined radiomics model.Results107 radiomic features were extracted from EUS and only those with nonzero coefficients were kept. Among the six radiomic models, the support vector machine (SVM) model had the highest performance with AUCs of 0.853 in the training cohort and 0.755 in the test cohort. A peritumoral radiomic model was developed and assessed, achieving an AUC of 0.841 in the training and 0.785 in the test cohorts. The amalgamated model, incorporating intratumoral and peritumoral radiomic features, exhibited superior predictive accuracy in both the training (AUC=0.861) and test (AUC=0.822) cohorts. These findings were validated using the Delong test. The calibration and decision curve analyses (DCA) of the combined radiomic model displayed exceptional accuracy and provided the greatest net benefit for clinical decision-making when compared to other models. Finally, the nomogram also achieved an excellent performance.ConclusionsAn efficient and accurate EUS-based radiomic model incorporating intratumoral and peritumoral radiomic features was proposed and validated to accurately distinguish PNETs from pancreatic cancer. This research has the potential to offer novel perspectives on enhancing the clinical utility of EUS in the prediction of PNETs.

Inflammatory bowel disease (IBD) is intricately linked to neuropsychiatric comorbidities through gut-brain axis dysregulation. This study demonstrates that resveratrol (RSV), a natural polyphenol, alleviates DSS-induced colitis-associated anxiety and depression by reprogramming the microbiota─metabolite-barrier network. RSV (100 mg/kg/day) ameliorated DSS-associated anxiety-like behaviors in open field tests (peripheral zone time ↓12.6 • Dual-axis intervention: Resveratrol suppresses the Turicibacter-4-guanidinobutanoic acid-MyD88 pro-inflammatory cascade and activates the Muribaculum/Dubosiella-polyamine-ZO-1 repair axis, restoring gut-brain homeostasis. • First evidence linking microbial arginine metabolism reprogramming to coordinated gut-brain barrier repair and microglial M2 polarization. • Translational paradigm: Identifies microbiota-driven metabolic rewiring as a therapeutic strategy for IBD-associated neuropsychiatric disorders.