ObjectiveTo investigate the mechanism of Xiezhuo Jiedu prescription in the treatment of ulcerative colitis （UC） by inhibiting ferroptosis and alleviating intestinal mucosal injury based on the nuclear factor E₂ related factor 2/solute carrier family 7 member/glutathione peroxidase 4 （Nrf2/SLC7A11/GPX4） signaling pathway. MethodsA total of 60 male SD rats were divided into a normal group， a model group， high- and low-dose Xiezhuo Jiedu prescription groups （26.64 and 13.32 g·kg^-1， respectively）， a ferroptosis inhibitor group （Ferrostatin-1， 0.005 g·kg^-1）， and a mesalazine group （0.27 g·kg^-1）， with 10 rats in each group. A UC rat model was established by intrarectal administration of trinitrobenzene sulfonic acid （TNBS）-ethanol. The normal group and the model group were intragastrically administered normal saline. The other groups were given intragastric administration according to the corresponding dosage for 7 d. The general condition， disease activity index （DAI） score， colon length， and mucosal injury index （CDMI） score were observed in each group. The pathological changes of colon tissue in each group were observed by hematoxylin-eosin （HE） staining. The intestinal mucosa and mitochondrial morphology in each group were observed by transmission electron microscopy. The expression levels of Occludin， Claudin-1， mucin 2 （MUC2）， and E-cadherin in intestinal tissue were detected by immunofluorescence （IF）. Enzyme-linked immunosorbent assay （ELISA） was used to detect the expression levels of serum tumor necrosis factor-α （TNF-α）， interleukin-6 （IL-6）， and interleukin-10 （IL-10） in each group， and a lactic acid assay kit or ELISA was employed to detect the expression levels of reactive oxygen species （ROS）， ferrous ions （Fe²⁺）， glutathione （GSH）， malondialdehyde （MDA）， 4-hydroxynonenal （4-HNE）， diamine oxidase （DAO）， and D-lactate （D-LA）. Real-time quantitative polymerase chain reaction （Real-time PCR） was applied to detect the mRNA expression levels of Nrf2， SLC7A11， GPX4， Occludin， Claudin-1， MUC2， and E-cadherin in each group， and Western blot was adopted to detect the protein expression levels of Nrf2， p-Nrf2， SLC7A11， and GPX4 in each group. ResultsCompared with the normal group， rats in the model group exhibited listlessness， sluggish response， and mucopurulent and bloody stools. The model group also showed significantly increased DAI score， colon length， CDMI score， and expression levels of TNF-α， IL-6， ROS， Fe²⁺， MDA， 4-HNE， DAO， and D-LA （P<0.01）. In addition， it presented significantly decreased IF values of Occludin， Claudin-1， MUC2， and E-cadherin and mRNA and protein expression levels of IL-10， GSH， Nrf2， p-Nrf2， SLC7A11， and GPX4 （P<0.01）. There were different degrees of improvement in each administration group after treatment， and the improvement was the most significant in the high-dose Xiezhuo Jiedu prescription group （P<0.01）. ConclusionXiezhuo Jiedu prescription may alleviate intestinal mucosal injury by inhibiting ferroptosis of intestinal epithelial cells via regulating the Nrf2/SLC7A11/GPX4 signaling pathway， thereby exhibiting efficacy in the treatment of UC.

OBJECTIVE To study the improvement effect and mechanism of Tripterygium wilfordii multiglucoside （TWM） on nephrotic syndrome in rats. METHODS The nephrotic syndrome model was established by intravenous injection of adriamycin via the tail vein. The modeling rats were randomly divided into the model group （distilled water）， prednisone group （10 mg/kg）， and TWM high- and low-dose groups （10 and 5 mg/kg， respectively）. Additionally， blank group （distilled water） without model induction was established. Each group consisted of 9 rats. Rats in each group were administered the corresponding drugs or distilled water by gavage， once a day， for 6 consecutive weeks. The histopathological morphology of kidney tissues in rats was observed； the levels of 24-hour urinary protein （24 h-UTP） and serum biochemical indicators [albumin （ALB）， blood urea nitrogen （BUN）， serum creatinine （SCr）， cholesterol （CHOL）， and triglyceride （TG）] in rats were determined； the levels of oxidative stress indicators [superoxide dismutase （SOD）， malondialdehyde （MDA）] in kidney tissue of rats were determined； expressions of hypoxia-inducible factor-1α （HIF-1α）/microRNA-155-5p （miR-155-5p）/nuclear factor erythriod 2- related factor 2 （Nrf2） signaling pathway-related mRNA and protein in the renal tissues of rats were detected. RESULTS Compared with the blank group， the rats in the model group exhibited disordered renal tissue structure， with a small amount of glomerular necrosis and edema of the renal tubular epithelial cells. 24 h-UTP， serum levels of SCr， BUN， CHOL and TG， MDA content， mRNA and protein expressions of HIF-1α and Keap1 as well as the expression of miR-155-5p in renal tissues were increased significantly （ P ＜0.05）. Serum level of ALB， SOD level in renal tissue as well as mRNA and protein expressions of Nrf2 were decreased significantly （ P ＜0.05）. Compared with the model group， TWM high-dose and low-dose groups exhibited significant improvements in renal injury， with notable reversals in the levels of the above quantitative indicators （ P ＜0.05）. CONCLUSIONS TWM can alleviate oxidative stress-induced damage and thereby improve nephrotic syndrome in rats by regulating the HIF-1α/miR-155-5p/Nrf2 signaling pathway.

Objective To explore the impact of number of positive regional lymph nodes (nPRLN) in N1 stage on the prognosis of non-small cell lung cancer (NSCLC) patients. Methods Patients with TxN1M0 stage NSCLC who underwent lobectomy and mediastinal lymph node dissection from 2010 to 2015 were screened from SEER database (17 Regs, 2022nov sub). The optimal cutoff value of nPRLN was determined using X-tile software, and patients were divided into 2 groups according to the cutoff value: a nPRLN≤optimal cutoff group and a nPRLN>optimal cutoff group. The influence of confounding factors was minimized by propensity score matching (PSM) at a ratio of 1 : 1. Kaplan-Meier curves and Cox proportional hazards models were used to evaluate overall survival (OS) and lung cancer-specific survival (LCSS) of patients. Results A total of 1316 patients with TxN1M0 stage NSCLC were included, including 662 males and 654 females, with a median age of 67 (60, 73) years. The optimal cutoff value of nPRLN was 3, with 1165 patients in the nPRLN≤3 group and 151 patients in the nPRLN>3 group. After PSM, there were 138 patients in each group. Regardless of before or after PSM, OS and LCSS of patients in the nPRLN≤3 group were superior to those in the nPRLN>3 group (P<0.001). N1 stage nPRLN>3 was an independent prognostic risk factor for OS [HR=1.52, 95%CI (1.22, 1.89), P<0.001] and LCSS [HR=1.72, 95%CI (1.36, 2.18), P<0.001]. Conclusion N1 stage nPRLN>3 is an independent prognostic risk factor for NSCLC patients in TxN1M0 stage, which may provide new evidence for future revision of TNM staging N1 stage subclassification.

[摘 要] 目的：探究环状核酸适配体-药物偶联物（ApDC）联合抗CD19嵌合抗原受体T/NK（CAR19-T/NK）细胞对Nalm6白血病细胞杀伤效果的增强作用。方法： 用数据库分析寻找白血病细胞的标志物蛋白酪氨酸激酶7（PTK7），设计合成靶向PTK7的核酸适配体sgc8，进一步合成两种ApDC：核酸适配体偶联吉西他滨（sgc8-GEM）和环状核酸适配体偶联吉西他滨（C-sgc8-GEM），琼脂糖凝胶电泳分析其稳定性。用CD19-CAR过表达慢病毒转染人外周血T/NK细胞构建CAR19-T和CAR19-NK细胞，流式细胞术检测sgc8与Nalm6细胞的结合力，CCK-8法和流式细胞术检测sgc8-GEM、C-sgc8-GEM，以及其分别与CAR19-T和CAR19-NK细胞联合应用对Nalm6细胞的杀伤能力。结果： PTK7 mRNA在白血病细胞中呈高表达（P < 0. 01）；C-sgc8-GEM比sgc8-GEM稳定性更高，与GEM比较，sgc8-GEM更能有效杀伤Nalm6细胞（P < 0.001）；C-sgc8-GEM或sgc8-GEM分别与CAR19-T/NK细胞联合应用比单独用C-sgc8-GEM、sgc8-GEM、CAR19-T/NK细胞对Nalm6细胞的毒性均更强（P < 0.001）。结论：成功建立环状ApDC与CAR19-T/NK细胞联合应用法，通过双靶点有效提高了对Nalm6细胞的毒性，为急性淋巴细胞白血病提供了新的治疗思路。

ObjectiveCleft palate (CP) is a common congenital deformity often associated with velopharyngeal insufficiency (VPI), which disrupts the physiological coupling between respiration and speech. Conventional clinical assessments, such as nasometry and spirometry, provide limited static data and fail to visualize the dynamic spatiotemporal distribution of lung ventilation during phonation. This study introduces spatiotemporal electrical impedance tomography (ST-EIT) to evaluate speech-respiratory functional features in CP patients compared to normal controls (NC). The aim is to characterize multi-domain respiratory patterns and to validate an interpretable machine learning framework for providing objective, quantitative evidence for clinical assessment. MethodsSeventy-five participants were enrolled in this study, comprising 37 patients with surgically repaired CP and 38 healthy volunteers matched for age, gender, and body mass index (BMI). All subjects performed standardized sustained phonation tasks while undergoing synchronous monitoring with a 16-electrode EIT system and a pneumotachograph. A comprehensive feature engineering pipeline was developed to extract physiological parameters across 3 complementary domains. (1) Temporal domain: including inspiratory/expiratory phase duration (tPhase), time constants (Tau), and inspiratory-to-expiratory time ratios (TI/TE); (2) airflow domain: comprising mean flow, peak flow, and instantaneous flow at 25%, 50%, and 75% of tidal volume; and (3) spatial domain: quantifying global and regional tidal impedance variation (TIV), global inhomogeneity (GI), and center of ventilation (CoV). Extreme Gradient Boosting (XGBoost) classifiers were trained using 5 distinct data sources (Spirometry, Nasometry, Inspiratory-EIT, Expiratory-EIT, and fused ST-EIT). Model performance was rigorously evaluated via stratified 5-fold cross-validation, and Shapley additive explanations (SHAP) were employed to quantify global and local feature contributions. ResultsThe CP group exhibited a distinct respiratory phenotype compared to controls. In the temporal domain, CP patients showed significantly shorter inspiratory (1.60 s vs.1.85 s, P<0.001) and expiratory phase durations (2.45 s vs. 3.95 s, P<0.001), indicating a rapid, shallow breathing rhythm. In the airflow domain, while inspiratory flows were comparable, the CP group demonstrated significantly elevated mean and peak flows during the expiratory phase (P<0.001), reflecting compensatory respiratory effort. Spatially, CP patients presented significant ventilation redistribution, characterized by higher regional TIV in the right-anterior (ROI1) and left-posterior (ROI4) quadrants, but lower TIV in the left-anterior (ROI2) quadrant. In terms of diagnostic accuracy, the multi-modal ST-EIT model achieved the highest performance (AUC: 0.915±0.012, Accuracy: 0.843±0.019, F1-score: 0.872±0.017), substantially outperforming models based on spirometry (AUC: 0.721) or nasometry (AUC: 0.625) alone. Interpretability analysis revealed that spatial domain features were the most critical, contributing 53.4% to the model’s decision-making, followed by temporal (25.0%) and airflow (21.6%) features. ConclusionST-EIT successfully captures the temporal, airflow, and spatial deviations in CP speech respiration that are undetectable by conventional methods—specifically, rapid phase transitions, hyperdynamic expiratory airflow, and regional ventilation heterogeneity. This study validates ST-EIT as a robust, non-invasive, and radiation-free tool for characterizing speech-respiratory dysfunction, offering high clinical value for bedside screening, rehabilitation planning, and longitudinal monitoring of patients with cleft palate.

ObjectiveCleft palate (CP) is a common congenital deformity often associated with velopharyngeal insufficiency (VPI), which disrupts the physiological coupling between respiration and speech. Conventional clinical assessments, such as nasometry and spirometry, provide limited static data and fail to visualize the dynamic spatiotemporal distribution of lung ventilation during phonation. This study introduces spatiotemporal electrical impedance tomography (ST-EIT) to evaluate speech-respiratory functional features in CP patients compared to normal controls (NC). The aim is to characterize multi-domain respiratory patterns and to validate an interpretable machine learning framework for providing objective, quantitative evidence for clinical assessment. MethodsSeventy-five participants were enrolled in this study, comprising 37 patients with surgically repaired CP and 38 healthy volunteers matched for age, gender, and body mass index (BMI). All subjects performed standardized sustained phonation tasks while undergoing synchronous monitoring with a 16-electrode EIT system and a pneumotachograph. A comprehensive feature engineering pipeline was developed to extract physiological parameters across 3 complementary domains. (1) Temporal domain: including inspiratory/expiratory phase duration (tPhase), time constants (Tau), and inspiratory-to-expiratory time ratios (TI/TE); (2) airflow domain: comprising mean flow, peak flow, and instantaneous flow at 25%, 50%, and 75% of tidal volume; and (3) spatial domain: quantifying global and regional tidal impedance variation (TIV), global inhomogeneity (GI), and center of ventilation (CoV). Extreme Gradient Boosting (XGBoost) classifiers were trained using 5 distinct data sources (Spirometry, Nasometry, Inspiratory-EIT, Expiratory-EIT, and fused ST-EIT). Model performance was rigorously evaluated via stratified 5-fold cross-validation, and Shapley additive explanations (SHAP) were employed to quantify global and local feature contributions. ResultsThe CP group exhibited a distinct respiratory phenotype compared to controls. In the temporal domain, CP patients showed significantly shorter inspiratory (1.60 s vs.1.85 s, P<0.001) and expiratory phase durations (2.45 s vs. 3.95 s, P<0.001), indicating a rapid, shallow breathing rhythm. In the airflow domain, while inspiratory flows were comparable, the CP group demonstrated significantly elevated mean and peak flows during the expiratory phase (P<0.001), reflecting compensatory respiratory effort. Spatially, CP patients presented significant ventilation redistribution, characterized by higher regional TIV in the right-anterior (ROI1) and left-posterior (ROI4) quadrants, but lower TIV in the left-anterior (ROI2) quadrant. In terms of diagnostic accuracy, the multi-modal ST-EIT model achieved the highest performance (AUC: 0.915±0.012, Accuracy: 0.843±0.019, F1-score: 0.872±0.017), substantially outperforming models based on spirometry (AUC: 0.721) or nasometry (AUC: 0.625) alone. Interpretability analysis revealed that spatial domain features were the most critical, contributing 53.4% to the model’s decision-making, followed by temporal (25.0%) and airflow (21.6%) features. ConclusionST-EIT successfully captures the temporal, airflow, and spatial deviations in CP speech respiration that are undetectable by conventional methods—specifically, rapid phase transitions, hyperdynamic expiratory airflow, and regional ventilation heterogeneity. This study validates ST-EIT as a robust, non-invasive, and radiation-free tool for characterizing speech-respiratory dysfunction, offering high clinical value for bedside screening, rehabilitation planning, and longitudinal monitoring of patients with cleft palate.

Objective To evaluate the safety and efficacy of a self-developed micro-normothermic machine perfusion (NMP) system (micro-perfusion device) for preserving isolated porcine limbs. Methods Five healthy Landrace pigs were selected, and their left and right forelimbs were randomly divided into the NMP group and static cold storage (SCS) group. The NMP group was perfused with the self-developed micro-perfusion device and polymerized hemoglobin perfusate for 32 hours at normothermia, while the SCS group was preserved at 4 ℃. Hemodynamic parameters such as perfusion pressure and flow were monitored. The pH value, partial pressure of oxygen (PO₂), lactic acid (Lac), creatine kinase (CK) and lactate dehydrogenase (LDH) in the perfusate were measured. Hematoxylin-eosin staining was used to assess the muscle tissue structure, terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling was employed to evaluate muscle cell apoptosis, and immunohistochemistry staining was applied to detect the expressions of tumor necrosis factor (TNF)-α and interleukin (IL)-6. A mixed-effects model was used to analyze the effects of time and treatment methods on tissue structure, cell apoptosis and inflammatory factors. Results The device could stably maintain a perfusion pressure of (69±15) mmHg and a flow rate of (117±42) mL/min. The pH value and electrolytes of the perfusate were generally stable, with PO₂ maintained at a high level. Lac was maintained at 5.38(3.81, 6.45) mmol/L, while CK and LDH increased over time. After 32 hours of perfusion in the NMP group, both the myocyte spacing and apoptosis rate were better than those in the SCS group. Mixed-effects model analysis showed that there were statistically significant differences in the effects of NMP treatment and SCS treatment on myocyte spacing and apoptosis rate per unit time (both P < 0.05). There were no statistically significant differences in TNF-α and IL-6 between the two groups, and mixed-effects model analysis showed no statistically significant differences in the effects of NMP treatment and SCS treatment on TNF-α and IL-6 per unit time (both P > 0.05). Conclusions The micro-perfusion device used in this study may achieve 32-hour normothermic preservation in a porcine limb amputation model, maintain basic metabolism and ionic homeostasis, reduce muscle structural damage and cell apoptosis without inducing additional inflammatory responses. This technology is expected to significantly extend the time window for replantation of amputated limbs in disaster rescue and long-distance transportation, providing an important technical basis for clinical translation and subsequent replantation research.

Background: Visceral white adipose tissue (vWAT) inflammation is a critical pathology of obesity-caused heart damage and is closely associated with adipocyte endoplasmic reticulum (ER) dysfunction. Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) has been identified as an adipokine with anti-vWAT inflammatory effects. However, it remains unclear whether Serpina3c deficiency promotion of vWAT inflammation involves adipocyte ER dysfunction and whether it further contributes to heart damage in obesity. Methods: Wild type and Serpina3c knockout (Serpina3c–/–) mice were fed a high-fat diet (HFD) for 12 weeks. An adeno-associated virus (AAV) was injected locally into epididymal white adipose tissue (eWAT) of Serpina3c–/– mice to induce eWAT-adipocyte- specific overexpression of Serpina3c (AAV-Serpina3c) or knockdown of hypoxia-inducible factor 1α (AAV-shHIF1α). In vitro experiments were performed in 3T3-L1 adipocytes. Results: Serpina3c–/– mice exhibited more severe eWAT, serum and heart inflammation after HFD feeding. Consistently, these adverse phenotypes were mitigated in AAV-Serpina3c and AAV-shHIF1α mice. Mechanistically, ER oxidoreductase 1α (Ero1α) and protein disulfide isomerase (PDI) family members PDIA3 and PDIA4 were found to be target genes of HIF1α. In the obese mice, Serpina3c deficiency caused adipocyte more hypertrophy, and activated HIF1α-Ero1α/PDI mediated ER overoxidation and ER stress in eWAT. Subsequently, this led to increased adipocyte apoptosis and chemokine production and decreased adiponectin expression, which promoted macrophage infiltration and M1 polarization in eWAT, thus exacerbating eWAT inflammation and ultimately facilitating serum and distal heart inflammation. Conclusion These findings indicate that Serpina3c is a significant regulator of adipocyte ER redox homeostasis, thus highlighting Serpina3c as a potential therapeutic target for obesity-related eWAT inflammation and heart damage.

ObjectiveIn traditional Chinese medicine (TCM), the foundational doctrine that the eyes reflect the essence of the internal viscera establishes ocular observation as a cornerstone of diagnostic practice. Specifically, the morphological characteristics and coloration variations of the scleral microvasculature serve as critical clinical indicators for assessing the dynamic balance of Qi and Blood, as well as the pathological status of internal organs. Historically, however, TCM eye diagnosis has relied predominantly on the subjective clinical experience and visual acuity of individual practitioners, leading to inherent challenges in standardization and reproducibility. While automated computer-aided diagnostic systems offer a promising solution, existing vessel segmentation algorithms encounter significant domain-specific bottlenecks when applied to scleral imagery. These challenges primarily stem from the highly reflective and moist nature of the ocular surface, which generates severe reflective interference. Furthermore, the inherent low contrast of fine capillary networks against complex background textures, compounded by non-uniform illumination, frequently results in high false-positive rates, misdetections, and severe vessel fragmentation. To address these critical limitations and advance the objective quantification of TCM diagnostics, this paper proposes a novel, highly robust sclera vessel segmentation framework that innovatively integrates Frangi-Sato dual-filter adaptive enhancement with pixel-level reflection detection. MethodsThe proposed methodology systematically addresses the segmentation pipeline through three synergistic stages. First, to overcome the structural limitations of single-filter approaches, a multi-scale weighted fusion strategy is meticulously designed to harness the complementary extraction capabilities of both Frangi and Sato filters. This adaptive enhancement optimally balances the preservation of main vessel trunk continuity with the heightened sensitivity required for delineating delicate, low-contrast peripheral capillaries. Second, to tackle the persistent issue of reflective highlights, a sophisticated multi-feature synergistic reflection detection module is introduced. By jointly analyzing local information entropy, gradient field variations, and intensity statistical distributions, this module achieves precise, pixel-level identification and elimination of reflective artifacts without compromising the underlying vascular structures. Finally, a dual-level adaptive thresholding strategy, featuring an innovative “core protection” mechanism, is implemented. This critical step effectively suppresses complex background noise while rigorously preserving the structural and topological integrity of the intricate vessel network, preventing the structural breaks often seen in conventional binarization methods. ResultsThe efficacy of the proposed framework was rigorously evaluated using both self-constructed clinical datasets specifically acquired for TCM research and standardized public datasets. Extensive experimental results demonstrate that the proposed method consistently outperforms state-of-the-art traditional approaches and contemporary deep learning models. Specifically, the proposed method achieves a Dice similarity coefficient of approximately 0.71 on the private clinical dataset, and secures the best performance across the majority of quantitative metrics on both datasets. Notably, the framework exhibits exceptional robustness and generalization capabilities in highly challenging scenarios characterized by intense reflective interference, low signal-to-noise ratios, and cross-domain image variations. ConclusionThis study successfully realizes the high-integrity, automated segmentation of scleral vessel networks under complex clinical imaging conditions. By overcoming the fundamental algorithmic challenges of reflection interference and micro-vessel loss, the proposed methodology provides potential support for the digitization, objective standardization, and intelligent advancement of modern TCM eye diagnosis systems.

ObjectiveIn traditional Chinese medicine (TCM), the foundational doctrine that the eyes reflect the essence of the internal viscera establishes ocular observation as a cornerstone of diagnostic practice. Specifically, the morphological characteristics and coloration variations of the scleral microvasculature serve as critical clinical indicators for assessing the dynamic balance of Qi and Blood, as well as the pathological status of internal organs. Historically, however, TCM eye diagnosis has relied predominantly on the subjective clinical experience and visual acuity of individual practitioners, leading to inherent challenges in standardization and reproducibility. While automated computer-aided diagnostic systems offer a promising solution, existing vessel segmentation algorithms encounter significant domain-specific bottlenecks when applied to scleral imagery. These challenges primarily stem from the highly reflective and moist nature of the ocular surface, which generates severe reflective interference. Furthermore, the inherent low contrast of fine capillary networks against complex background textures, compounded by non-uniform illumination, frequently results in high false-positive rates, misdetections, and severe vessel fragmentation. To address these critical limitations and advance the objective quantification of TCM diagnostics, this paper proposes a novel, highly robust sclera vessel segmentation framework that innovatively integrates Frangi-Sato dual-filter adaptive enhancement with pixel-level reflection detection. MethodsThe proposed methodology systematically addresses the segmentation pipeline through three synergistic stages. First, to overcome the structural limitations of single-filter approaches, a multi-scale weighted fusion strategy is meticulously designed to harness the complementary extraction capabilities of both Frangi and Sato filters. This adaptive enhancement optimally balances the preservation of main vessel trunk continuity with the heightened sensitivity required for delineating delicate, low-contrast peripheral capillaries. Second, to tackle the persistent issue of reflective highlights, a sophisticated multi-feature synergistic reflection detection module is introduced. By jointly analyzing local information entropy, gradient field variations, and intensity statistical distributions, this module achieves precise, pixel-level identification and elimination of reflective artifacts without compromising the underlying vascular structures. Finally, a dual-level adaptive thresholding strategy, featuring an innovative “core protection” mechanism, is implemented. This critical step effectively suppresses complex background noise while rigorously preserving the structural and topological integrity of the intricate vessel network, preventing the structural breaks often seen in conventional binarization methods. ResultsThe efficacy of the proposed framework was rigorously evaluated using both self-constructed clinical datasets specifically acquired for TCM research and standardized public datasets. Extensive experimental results demonstrate that the proposed method consistently outperforms state-of-the-art traditional approaches and contemporary deep learning models. Specifically, the proposed method achieves a Dice similarity coefficient of approximately 0.71 on the private clinical dataset, and secures the best performance across the majority of quantitative metrics on both datasets. Notably, the framework exhibits exceptional robustness and generalization capabilities in highly challenging scenarios characterized by intense reflective interference, low signal-to-noise ratios, and cross-domain image variations. ConclusionThis study successfully realizes the high-integrity, automated segmentation of scleral vessel networks under complex clinical imaging conditions. By overcoming the fundamental algorithmic challenges of reflection interference and micro-vessel loss, the proposed methodology provides potential support for the digitization, objective standardization, and intelligent advancement of modern TCM eye diagnosis systems.