Objective: To develop a dual-branch deep learning framework for accurate multi-label classification of fundus diseases, addressing the key limitations of insufficient complementary feature extraction and inadequate cross-modal feature fusion in existing automated diagnostic methods. Methods: The fundus multi-label classification dataset with 12 disease categories (FMLC-12) dataset was constructed by integrating complementary samples from Ocular Disease Intelligent Recognition (ODIR) and Retinal Fundus Multi-Disease Image Dataset (RFMiD), yielding 6 936 fundus images across 12 retinal pathology categories, and the framework was validated on both FMLC-12 and ODIR. Inspired by the holistic multi-regional assessment principle of the Five Wheels theory in traditional Chinese medicine (TCM) ophthalmology, the dual-branch multi-label network (DBMNet) was developed as a novel framework integrating complementary visual feature extraction with pathological correlation modeling. The architecture employed a TransNeXt backbone within a dual-branch design: one branch processed red-green-blue (RGB) images to capture color-dependent features, such as vascular patterns and lesion morphology, while the other processed grayscale-converted images to enhance subtle textural details and contrast variations. A feature interaction module (FIM) effectively integrated the multi-scale features from both branches. Comprehensive ablation studies were conducted to evaluate the contributions of the dual-branch architecture and the FIM. The performance of DBMNet was compared against four state-of-the-art methods, including EfficientNet Ensemble, transfer learning-based convolutional neural network (CNN), BFENet, and EyeDeep-Net, using mean average precision (mAP), F1-score, and Cohen's kappa coefficient. Results: The dual-branch architecture improved mAP by 15.44 percentage points over the single-branch TransNeXt baseline, increasing from 34.41% to 44.24%, and the addition of FIM further boosted mAP to 49.85%. On FMLC-12, DBMNet achieved an mAP of 49.85%, a Cohen’s kappa coefficient of 62.14%, and an F1-score of 70.21%. Compared with BFENet (mAP: 45.42%, kappa: 46.64%, F1-score: 71.34%), DBMNet outperformed it by 4.43 percentage points in mAP and 15.50 percentage points in kappa, while BFENet achieved a marginally higher F1-score. On ODIR, DBMNet achieved an F1-score of 85.50%, comparable to state-of-the-art methods. Conclusion DBMNet effectively integrates RGB and grayscale visual modalities through a dual-branch architecture, significantly improving multi-label fundus disease classification. The framework not only addresses the issue of insufficient feature fusion in existing methods but also demonstrates outstanding performance in balancing detection across both common and rare diseases, providing a promising and clinically applicable pathway for standardized, intelligent fundus disease classification.

ObjectiveBased on the theory of "spleen governing transportation and transformation"， this study investigates the efficacy of Atractylodis Macrocephalae Rhizoma processed with Aurantii Fructus Immaturus juice（AMR-AFI） in improving slow-transit constipation（STC）， as well as the synergistic regulatory mechanism involving the microbiota-metabolism axis， thereby elucidating the scientific basis of its processing theory. MethodsAnimals were randomly divided into the control group， model group， positive drug（mosapride） group（3 mg·kg^-1）， and low-， medium-， and high-dose groups of AMR-AFI（3.9， 7.8， 15.6 g·kg^-1）. Except for the control group， the remaining five groups were induced with STC using loperamide hydrochloride. Following modeling， interventions were administered. All groups received continuous administration for 15 d， during which fecal samples， colon tissue， and serum were collected. Constipation improvement was assessed by measuring fecal moisture content and small intestinal propulsion rate， histological morphology of colonic tissue was observed via hematoxylin-eosin（HE） staining， and the levels of interleukin（IL）-6， tumor necrosis factor（TNF）-α， and IL-2 in serum were detected using enzyme-linked immunosorbent assay（ELISA）. Furthermore， the microbial community structure in mouse feces was analyzed by 16S rRNA sequencing， while transcriptomic sequencing was employed to screen differentially expressed genes in colonic tissue， followed by gene ontology（GO） and Kyoto Encyclopedia of Genes and Genomes（KEGG） enrichment analyses. Finally， Spearman correlation analysis was conducted to explore the association between differential microbiota and differential genes. ResultsCompared with the control group， the intestinal propulsion rate and fecal moisture content in the model group were significantly decreased（P<0.01）， while serum levels of IL-6， TNF-α， and IL-2 were significantly elevated（P<0.01）. HE staining showed damage and shedding of colonic mucosal epithelial cells， along with a reduction in goblet cells in the model group. In comparison with the model group， all treatment groups improved the pathological state of the colonic mucosa to varying degrees and reduced serum levels of IL-6， TNF-α， and IL-2（P<0.01）. Among these， the high-dose group of AMR-AFI significantly increased the intestinal propulsion rate and fecal moisture content of rats（P<0.05， P<0.01）. Further transcriptomic analysis revealed that a total of 104 differentially expressed genes were identified from comparisons between the model group and the control group， as well as between the model group and the high-dose group of AMR-AFI. These genes were mainly enriched in pathways closely related to STC pathogenesis， such as arachidonic acid metabolism and aldosterone-regulated sodium reabsorption. 16S rRNA sequencing results indicated that AMR-AFI reversed the structural imbalance of the gut microbiota in model mice， increased species richness， downregulated the relative abundance of pro-inflammatory bacteria such as Parasutterella， and enriched beneficial and butyrate-producing bacteria， including Lachnospiraceae_NK4A136_group， Ruminococcaceae， and Lachnospiraceae. Spearman correlation analysis further showed that the beneficial bacteria enriched in the AMR-AFI group were negatively correlated with genes involved in the arachidonic acid metabolic pathway and positively correlated with genes in the aldosterone-regulated sodium reabsorption pathway. In contrast， pro-inflammatory bacteria in the model group exhibited the opposite correlation trends. ConclusionAMR-AFI can effectively exert synergistic therapeutic effects on STC by regulating intestinal microbiota， arachidonic acid-mediated inflammatory metabolism， and aldosterone-regulated water-salt balance pathways.

ObjectiveBased on the theory of "spleen governing transportation and transformation"， this study investigates the efficacy of Atractylodis Macrocephalae Rhizoma processed with Aurantii Fructus Immaturus juice（AMR-AFI） in improving slow-transit constipation（STC）， as well as the synergistic regulatory mechanism involving the microbiota-metabolism axis， thereby elucidating the scientific basis of its processing theory. MethodsAnimals were randomly divided into the control group， model group， positive drug（mosapride） group（3 mg·kg^-1）， and low-， medium-， and high-dose groups of AMR-AFI（3.9， 7.8， 15.6 g·kg^-1）. Except for the control group， the remaining five groups were induced with STC using loperamide hydrochloride. Following modeling， interventions were administered. All groups received continuous administration for 15 d， during which fecal samples， colon tissue， and serum were collected. Constipation improvement was assessed by measuring fecal moisture content and small intestinal propulsion rate， histological morphology of colonic tissue was observed via hematoxylin-eosin（HE） staining， and the levels of interleukin（IL）-6， tumor necrosis factor（TNF）-α， and IL-2 in serum were detected using enzyme-linked immunosorbent assay（ELISA）. Furthermore， the microbial community structure in mouse feces was analyzed by 16S rRNA sequencing， while transcriptomic sequencing was employed to screen differentially expressed genes in colonic tissue， followed by gene ontology（GO） and Kyoto Encyclopedia of Genes and Genomes（KEGG） enrichment analyses. Finally， Spearman correlation analysis was conducted to explore the association between differential microbiota and differential genes. ResultsCompared with the control group， the intestinal propulsion rate and fecal moisture content in the model group were significantly decreased（P<0.01）， while serum levels of IL-6， TNF-α， and IL-2 were significantly elevated（P<0.01）. HE staining showed damage and shedding of colonic mucosal epithelial cells， along with a reduction in goblet cells in the model group. In comparison with the model group， all treatment groups improved the pathological state of the colonic mucosa to varying degrees and reduced serum levels of IL-6， TNF-α， and IL-2（P<0.01）. Among these， the high-dose group of AMR-AFI significantly increased the intestinal propulsion rate and fecal moisture content of rats（P<0.05， P<0.01）. Further transcriptomic analysis revealed that a total of 104 differentially expressed genes were identified from comparisons between the model group and the control group， as well as between the model group and the high-dose group of AMR-AFI. These genes were mainly enriched in pathways closely related to STC pathogenesis， such as arachidonic acid metabolism and aldosterone-regulated sodium reabsorption. 16S rRNA sequencing results indicated that AMR-AFI reversed the structural imbalance of the gut microbiota in model mice， increased species richness， downregulated the relative abundance of pro-inflammatory bacteria such as Parasutterella， and enriched beneficial and butyrate-producing bacteria， including Lachnospiraceae_NK4A136_group， Ruminococcaceae， and Lachnospiraceae. Spearman correlation analysis further showed that the beneficial bacteria enriched in the AMR-AFI group were negatively correlated with genes involved in the arachidonic acid metabolic pathway and positively correlated with genes in the aldosterone-regulated sodium reabsorption pathway. In contrast， pro-inflammatory bacteria in the model group exhibited the opposite correlation trends. ConclusionAMR-AFI can effectively exert synergistic therapeutic effects on STC by regulating intestinal microbiota， arachidonic acid-mediated inflammatory metabolism， and aldosterone-regulated water-salt balance pathways.

ObjectiveTo investigate the therapeutic effects and mechanisms of modified Huangqi Jianzhong decoction （HQJZ） on gastric precancerous conditions （GPC）. MethodsIn the cell experiment， human gastric mucosal epithelial cells underwent malignant transformation induced by N-methyl-N′-nitro-N-nitrosoguanidine （MNNG） for the modeling of GPC （MC cells）. The cells were allocated into four groups： control ， model， low-dose HQJZ （HQJZ-L）， and high-dose HQJZ （HQJZ-H）. The control and model groups were cultured with the complete medium， while HQJZ-L and HQJZ-H groups received additional interventions with HQJZ at low （0.5 g·L^-1） and high （1.0 g·L^-1） doses， respectively. Cell counting kit-8 （CCK-8） assay was used to evaluate cytotoxicity， Transwell assay to assess cell invasion， Annexin V-FITC/PI staining to detect apoptosis， immunofluorescence assay to analyze reactive oxygen species （ROS） expression and mitochondrial autophagy， and Western blot to verify expression of proteins in key pathways. In the animal experiment， the GPC model was established in healthy BALB/c mice through MNNG induction. Twenty-four mice were allocated into four groups： control， model， HQJZ-L， and HQJZ-H. Control and model groups received normal saline （10 mL·kg^-1·d^-1） orally， while HQJZ-L and HQJZ-H groups were administrated with low-dose （6.24 g·kg^-1·d^-1） and high-dose （12.48 g·kg^-1·d^-1） HQJZ， respectively. After treatment， hematoxylin‑eosin （HE） staining and AB-PAS staining were performed to observe histopathological changes in the gastric tissue. Immunofluorescence assay was used to detect reactive oxygen species （ROS） and microtubule-associated protein 1 light chain 3 （LC3） levels in the gastric mucosa， TdT-mediated dUTP nick-end labeling （TUNEL） staining to assess apoptosis rates， and Western blotting and immunohistochemistry （IHC） to analyze the expression levels of Ki67， proliferating cell nuclear antigen （PCNA）， and foxhead box O3 （FoxO3）. ResultsCell viability assays showed that HQJZ dose-dependently reduced MC cell viability compared with the control group （P<0.05， P<0.01）. Transwell assays revealed that the model group exhibited enhanced cell invasion compared with the control group （P<0.05）. Compared with the model group， HQJZ treatment attenuated the cell invasion （P<0.05）. Gastric mucosal pathology in mice demonstrated that compared with the control group， the model group showed elevated HE and AB-PAS pathological scores （P<0.05）， while HQJZ treatment reduced these scores （P<0.05）. Transmission electron microscopy revealed increased mitochondrial number and volume in the model group compared with the control group. HQJZ treatment resulted in abnormal mitochondrial structure and significant alterations in rough endoplasmic reticulum morphology and distribution， presenting as dilated and hollow forms. Mitochondrial and apoptosis assessments indicated that compared with the control group， the model group exhibited enhanced Mito Tracker Green fluorescence （P<0.05）， no significant change in DCFH-DA fluorescence， Mito Tracker Red CMXRos fluorescence， ROS immunofluorescence， or malondialdehyde （MDA） level， increased GSH level （P<0.05）， enhanced LC3 fluorescence （P<0.05）， no significant change in apoptosis rate， and elevated ATP content in cells and mouse serum （P<0.05）. Compared with the model group， HQJZ treatment reduced Mito Tracker Green fluorescence （P<0.05）， increased DCFH-DA fluorescence， Mito Tracker Red fluorescence， MDA level， LC3 fluorescence， and apoptosis rate （P<0.05）， and decreased cellular ATP content （P<0.05）. The HQJZ-L group showed no significant change in ROS immunofluorescence or GSH level， whereas the HQJZ-H group demonstrated enhanced ROS immunofluorescence and glutathione （GSH） level （P<0.05）. Immunohistochemistry and Western blotting revealed that compared with the control group， the model group exhibited increased numbers of PCNA- and Ki67-positive cells （P<0.05） and elevated protein levels of FoxO3， sirtuin 1 （SIRT1）， and B-cell lymphoma 6 （Bcl-6） （P<0.05）. HQJZ treatment reduced the numbers of PCNA- and Ki67-positive cells （P<0.05） and lowered the protein levels of FoxO3， SIRT1， and Bcl-6 （P<0.05）. ConclusionHQJZ alleviates the progression of gastric precancerous lesions by regulating mitochondrial function via the FoxO3/ROS pathway and promoting apoptosis of GPC-malignant cells.

BackgroundRumination can play a certain degree role of psychological adjustment in cancer patients. Previous studies have focused on studying the level of rumination in a single type of cancer patient， but there is a lack of comprehensive investigation and influencing factor research on rumination levels in different types of cancer patients. ObjectiveTo explore the level of rumination and its influencing factors among middle-advanced cancer inpatients， so as to provide certain guidance for targeted psychological care in clinical practice. MethodsFrom January 2021 to December 2022， a systematic sampling technique was used to recruit 346 patients with TNM stage III or above tumors hospitalized in Nuclear Industry 416 Hospital， Chengdu Sixth People's Hospital and the First People's Hospital of Yibin. All individuals were assessed using Perceived Social Support Scale （PSSS） and the Chinese version of Event-Related Ruminant Inventory （C-ERRI）， and Logistic regression analysis was utilized to identify the influencing factors of rumination. ResultsIn terms of C-ERRI， patients scored （15.59±5.61） on intrusive rumination and （14.59±5.43） on deliberate rumination. Education levels of junior high school/high school/vocational training school （OR=0.817， P<0.01） and junior college and above （OR=0.579， P<0.05） were the protective factors of intrusive rumination， whereas annual personal incomes of <10 000 yuan （OR=4.918， P<0.01） or 10 000~50 000 yuan （OR=2.076， P<0.01） and low （OR=6.882， P<0.01） or middle （OR=3.114， P<0.01） level of social support were the risk factors of intrusive rumination. For deliberate rumination， education levels of junior high school/high school/vocational training school （OR=0.574， P<0.01） and junior college and above （OR=0.449， P<0.05） were the protective factors， and low （OR=1.391， P<0.01） or middle （OR=1.161， P<0.05） levels of social support were the risk factors. ConclusionThe level of intrusive rumination of inpatients with middle-advanced cancer is related to education level， economic status and social support， furthermore， the level of deliberate rumination is related to the educational level and social support. ［Funded by Medical Research Project of Chengdu Health Commission （ number， 2020119）］

Background: A simple superoposterior approach to thoracic transforaminal epidural injections (TFEIs) under ultrasonographic guidance was proposed to reduce zoster-associated pain (ZAP) involving multiple thoracic nerves and the likelihood of transitioning to postherpetic neuralgia (PHN). Methods: Patients were prospectively enrolled. Primary endpoints were the burden of illness (BOI) scores and epidural contrast spread. Secondary endpoints included number of needle insertion attempts, sensory blockade, hemodynamic changes, procedure time, radiation dose, adverse events, rescue analgesics, PHN incidence and EuroQoL 5-Dimension scores. Results: Thirty-five injections were performed in 27 patients. Median levels of cephalad-caudad epidural contrast spread were 3, 4, and 5 ml following injections of 2, 3, and 4 ml. Dorsal epidural spread was observed at levels 3, 4, and 5, whereas concurrent ventral spread was observed at levels 2, 3, and 4. BOI scores at 30–180 days significantly decreased (mean difference: −25.3, 95% CI [−57.4 to 6.6], P = 0.005), accounting for reduced rescue analgesic requirements and PHN occurrence and improved EuroQoL 5-Dimension scores. Median sensory blockade at 5 min post-procedure was at level 2, 3, and 4 after 2, 3, and 4 ml of therapeutic injectate. No significant hemodynamic changes were noted at 15 min post-injection. No serious adverse events were observed. Conclusions Spread of thoracic epidural contrast to all involved nerves was confirmed using this novel technique. Simplified needle placement reduced the technical difficulty and risk of complications. It might be a promising alternative approach for ZAP.

Background: A simple superoposterior approach to thoracic transforaminal epidural injections (TFEIs) under ultrasonographic guidance was proposed to reduce zoster-associated pain (ZAP) involving multiple thoracic nerves and the likelihood of transitioning to postherpetic neuralgia (PHN). Methods: Patients were prospectively enrolled. Primary endpoints were the burden of illness (BOI) scores and epidural contrast spread. Secondary endpoints included number of needle insertion attempts, sensory blockade, hemodynamic changes, procedure time, radiation dose, adverse events, rescue analgesics, PHN incidence and EuroQoL 5-Dimension scores. Results: Thirty-five injections were performed in 27 patients. Median levels of cephalad-caudad epidural contrast spread were 3, 4, and 5 ml following injections of 2, 3, and 4 ml. Dorsal epidural spread was observed at levels 3, 4, and 5, whereas concurrent ventral spread was observed at levels 2, 3, and 4. BOI scores at 30–180 days significantly decreased (mean difference: −25.3, 95% CI [−57.4 to 6.6], P = 0.005), accounting for reduced rescue analgesic requirements and PHN occurrence and improved EuroQoL 5-Dimension scores. Median sensory blockade at 5 min post-procedure was at level 2, 3, and 4 after 2, 3, and 4 ml of therapeutic injectate. No significant hemodynamic changes were noted at 15 min post-injection. No serious adverse events were observed. Conclusions Spread of thoracic epidural contrast to all involved nerves was confirmed using this novel technique. Simplified needle placement reduced the technical difficulty and risk of complications. It might be a promising alternative approach for ZAP.

Objective:To explore a factor analysis method for job training needs based on "P4 medicine", and to provide evidence-based support for optimizing continuing education systems for medical staff in military rest homes and sanatoriums.Methods:A cross-sectional study was conducted with 95 medical staff from military rest homes and sanatoriums within the support system who participated in job training at Xinqiao Hospital. A structured questionnaire was developed based on the "P4 medicine" framework. After two rounds of Delphi expert consultation, 46 core indicators were selected. A principal component analysis was used to extract common factors, and a model of hierarchical needs was constructed by combining varimax rotation and entropy weight methods.Results:Four common factors were extracted, accounting for 81.564% cumulative variance. The first factor "dynamic updating of geriatric medical knowledge" (31.83%) covered clinical core competencies in managing geriatric comorbidities and recognizing critical values. The second factor "emergency responsiveness and personalized care" (26.05%) focused on military medical regulations, emergency treatment protocols, and tailored interventions. The third factor "multidisciplinary collaboration and leadership development" (12.41%) emphasized team reorganization in combat-ready scenarios. The fourth factor "integration of intelligent technologies" (11.29%) reflected data-driven decision-making needs. Entropy weight analysis highlighted dynamic medical knowledge updates (e.g., emergency skills with a weight of 0.050) and AI applications with a weight of 0.019 among the top 10% high-weight indicators.Conclusions:The "P4 medicine" effectively delineates the multidimensional training needs of grassroots medical personnel. Its four-dimensional structure (predictive, normative, collaborative, and innovative) provides a theoretical framework for curriculum design. We recommend constructing tiered training modules prioritizing technology-enabled mechanisms to enhance the precision and sustainability of geriatric healthcare services.

Objective:To investigate the clinical subtypes of acute aortic dissection (AAD) through latent class analysis.Methods:This study was a multicenter retrospective cohort study. Patients with AAD admitted to five hospitals, including Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Henan Provincial People′s Hospital, Central China Fuwai Hospital of Zhengzhou University (Fuwai Central China Cardiovascular Hospital), the Third Affiliated Hospital of Xinxiang Medical University and the Second Affiliated Hospital of Chongqing Medical University, between August 2010 and December 2021 were enrolled. Based on clinical and biological characteristics, latent class analysis (models with 2 to 5 latent classes) was conducted to classify the enrolled patients. The optimal classification scheme was determined using model fitting evaluations, including log-likelihood (LL), entropy, Lo-Mendell-Rubin adjusted likelihood ratio test and so on. Clinical data of different subtypes were compared, and in-hospital mortality was analyzed across the entire population and among subgroups receiving different treatments.Results:A total of 2 689 AAD patients, aged 54 (46, 63) years were included, with 1 305 (48.5%) having DeBakey type Ⅰ, 156 (5.8%) type Ⅱ, and 1 228 (45. 7%) type Ⅲ dissections. The cohort comprised 2 134 (79.4%) males. The overall in-hospital mortality rate was 22.8% (613/2 689). Latent class analysis indicated that a two-class model was optimal (LL=147 413.242, entropy=0.812, and PLMRT<0.001). Patients were classified into two subtypes, named clinical subtype 1 and clinical subtype 2. Compared to clinical subtype 1, clinical subtype 2 had a higher proportion of females, was older, had more dissections involving the ascending aorta, and exhibited higher rates of organ dysfunction (elevated alanine aminotransferase and creatinine levels) and inflammatory response (neutrophilia) (all P<0.05). Clinical subtype 2 also showed higher in-hospital mortality compared to subtype 1 (26.3% (238/905) vs. 21.0% (375/1 784), P=0.002). Among patients undergoing surgical treatment, clinical subtype 2 had higher mortality than subtype 1 (40.1% (67/167) vs. 30.0% (101/337), P=0.027). However, no significant differences in mortality were observed between the two subtypes among patients receiving medical therapy or endovascular and hybrid procedures (all P>0.05). Conclusions:Comprehensive latent class analysis identifies two subtypes of AAD with distinct clinical characteristics and treatment responses. These findings provide new insights into individualized clinical decision-making and prognostic evaluation for AAD patients.

Objective:To explore the prevalence and risk factors of insomnia in Chinese Air Force servicemen deployed to highland areas.Methods:A total of 718 Air Force servicemen deployed to Qinghai-Tibetan plateau were recruited at May 2024.Sleep quality was assessed with the Pittsburgh Sleep Quality Index.Social-demograph-ic,military service,and psychological characteristics were measured with a self-administered general question-naire.Bivariable and multivariable logistic regressions were performed to identify independent risk factors.Missing data were handled by the multiple imputation.Results:The average sleep duration was(6.9±1.2)h and the aver-age PSQI score was(5.9±4.1).Totally 53.8％of participants experienced clinically significant insomnia.The multivariable analysis revealed that age≥35(aOR=4.07,95％CI=1.11-17.76),stressful event(aOR=3.27,95％CI=2.00-5.49),dysfunctional sleep beliefs and attitudes(aOR=2.59,95％CI=1.75-3.85),and caffeine product usage(aOR=1.69,95％CI=1.17-2.43)were risk factors for insomnia,while Tibetan-indigenous ethnic(aOR=0.44,95％CI=0.20-0.91),higher perceived social support(aOR=0.96,95％CI=0.96-0.99),and positive coping style(aOR=0.96,95％CI=0.93-0.99)were protective factors.Conclusion:Air force service-men deployed to highland areas have sufficient sleep time,but reduced sleep quality.Age,exposed to stress event during deployment,dysfunctional sleep beliefs and attitudes,and caffeine product usage are risk factors for insomni-a,while Tibetan-indigenous ethnic,higher perceived social support and positive coping style act as protective fac-tors.