ObjectiveTo address the challenges of unstructured classical Chinese expressions， nested entity relationships， and limited annotated data in famous traditional Chinese medicine（TCM） case records， this study proposes a joint relation extraction framework that integrates data augmentation and entity mapping， aiming to support the construction of TCM diagnostic knowledge graphs and clinical pattern mining. MethodsWe developed an annotation structure for entities and their relationships in TCM case texts and applied a data augmentation strategy by incorporating multiple ancient texts to expand the relation extraction dataset. A cascade binary tagging framework for relation triple extraction（CasRel） model for TCM semantics was designed， integrating a pre-trained bidirectional encoder representations from transformers（BERT） layer for classical TCM texts to enhance semantic representation， and using a head entity-relation-tail entity mapping mechanism to address entity nesting and relation overlapping issues. ResultsExperimental results showed that the CasRel model， combining data augmentation and entity mapping， outperformed the pipeline-based Bert-Radical-Lexicon（BRL）-bidirectional long short-term memory（BiLSTM）-Attention model. The overall precision， recall， and F₁-score across 12 relation types reached 65.73%， 64.03%， and 64.87%， which represent improvements of 14.26%， 7.98%， and 11.21% compared to the BRL-BiLSTM-Attention model， respectively. Notably， the F₁-score for tongue syndrome relations increased by 22.68%（69.32%）， and the prescription-syndrome relations performed the best with the F₁-score of 70.10%. ConclusionThe proposed framework significantly improves the semantic representation and complex dependencies in TCM texts， offering a reusable technical framework for structured mining of TCM case records. The constructed knowledge graph can support clinical syndrome differentiation， prescription optimization， and drug compatibility， providing a methodological reference for TCM artificial intelligence research.

Objective To construct prostate cancer(PCa)prediction models based on machine learning algorithms, so as to improve the accuracy of early diagnosis of PCa. Methods A retrospective analysis was performed on the clinical data of 504 patients who underwent prostate biopsy at our hospital during Jan. 2020 and Nov. 2024. Patients' age, body mass index(BMI), history of hypertension, diabetes and smoking, total prostate-specific antigen(tPSA), free prostate-specific antigen(fPSA), f/tPSA, prostate volume(PV), neutrophil count, lymphocyte count, neutrophil-to-lymphocyte ratio(NLR), Prostate Imaging Reporting and Data System(PI-RADS)score, digital rectal examination(DRE)results, and pathological findings were collected. The patients were divided into the training and testing sets at a ratio of 7:3. Ten early diagnosis prediction models of PCa were constructed using 10 supervised machine learning algorithms. Model performance was evaluated and validated using metrics including area under the receiver operating characteristic curve(AUC), accuracy, sensitivity, specificity, calibration curves, and decision curve analysis(DCA). SHAP analysis was used to interpret the models, and to clarify the importance of each feature and the basis for model decisions. Results All models showed good predictive value, with the gradient boosting machine(GBM)model performing the best(AUC=0.905, accuracy=84.1%, sensitivity=90.2%, specificity=80.0%). Calibration curves indicated good calibration and fitting of the GBM model, while DCA demonstrated favorable clinical net benefits. SHAP analysis identified the most significant features affecting PCa occurrence in descending order:tPSA, f/tPSA, PIRADS score, PV, age, and DRE. Conclusion The GBM model exhibits the optimal performance among the 10 models. The importance of features for predicting the occurrence of PCa, from the highest to the lowest, is as follows:tPSA, f/tPSA, PIRADS score, PV, age, and DRE.

BackgroundCurrently， the problem of depressed mood in college students is becoming more prominent. The experience of childhood maltreatment is a significant contributor to depression among college students. Although the association between the two has been confirmed， the specific psychosocial mechanisms underlying how childhood maltreatment affects college students' mental health remain insufficiently evidenced. ObjectiveTo explore the mediating role of emotion regulation difficulties in the relationship between childhood maltreatment and depression among college students， and to investigate the moderated effects of psychological resilience and family socioeconomic status， aiming to provide references for improving depressive symptoms in college students. MethodsOn 14 March 2024， a cluster sampling method was employed to recruit 751 college students from a university in Heilongjiang Province. Participants were assessed with Childhood Trauma Questionnaire （CTQ）， Difficulties in Emotion Regulation Scale （DERS）， Patients' Health Questionnaire Depression Scale-9 item （PHQ-9）， 10-item Connor-Davidson Resilience Scale （CD-RISC-10） and Family Socioeconomic Status Questionnaire. Pearson correlation analysis was adopted to examine the correlation between the scores of scales. Model 4 and model 7 in Process 4.2 were used to test the mediating effects of emotional regulation difficulties and the moderated effects of psychological resilience and family socioeconomic status. Results① A total of 712 （94.81%） valid questionnaires were collected. ② College students' CTQ score was positively correlated with DERS score and PHQ-9 score （r=0.296， 0.507， P<0.01）， and negatively correlated with CD-RISC-10 score and Family Socioeconomic Status Questionnaire score （r=-0.148， -0.229， P<0.01）. ③ The indirect effect value of difficulties in emotion regulation on the relationship between childhood maltreatment and depression was 0.091 （95% CI： 0.018~0.046）， accounting for 17.95% of the total effect. ④ The first half of the mediation model "childhood maltreatment → difficulties in emotion regulation → depression" （childhood maltreatment → difficulties in emotion regulation） was moderated by psychological resilience （β=-0.030， t=-6.147， 95% CI： -0.040~-0.020） and family socioeconomic status （β=-0.051， t=-3.929， 95% CI： -0.077~-0.026）. ConclusionChildhood maltreatment exerts both a direct effect on college students' depression and an indirect effect through emotion regulation difficulties. The childhood maltreatment → emotion regulation difficulties pathway in this mediation model is moderated by psychological resilience and family socioeconomic status. ［Funded by Qiqihar Medical University Graduate Student Innovation Fund Project （number， QYYCX2023-48）； Special Research Fund Project for Young Doctors of Qiqihar Academy of Medical Sciences （number， QMSI2021B-08）］

ObjectiveThis study aims to develop and validate a novel multimodal predictive model， termed criss-cross network（CCNet）-directed graph neural network（DGNN）（CGN）， for accurate assessment of pulmonary nodule progression in high-risk individuals for lung cancer， by integrating longitudinal chest computed tomography（CT） imaging with both traditional Chinese and western clinical evaluation data. MethodsA cohort of 4 432 patients with pulmonary nodules was retrospectively analyzed. A twin CCNet was employed to extract spatiotemporal representations from paired sequential CT scans. Structured clinical assessment and imaging-derived features were encoded via a multilayer perceptron， and a similarity-based alignment strategy was adopted to harmonize multimodal imaging features across temporal dimensions. Subsequently， a DGNN was constructed to integrate heterogeneous features， where nodes represented modality-specific embeddings and edges denoted inter-modal information flow. Finally， model optimization was performed using a joint loss function combining cross-entropy and cosine similarity loss， facilitating robust classification of nodule progression status. ResultsThe proposed CGN model demonstrated superior predictive performance on the held-out test set， achieving an area under the receiver operating characteristic curve（AUC） of 0.830， accuracy of 0.843， sensitivity of 0.657， specificity of 0.712， Cohen's Kappa of 0.417， and F₁ score of 0.544. Compared with unimodal baselines， the CGN model yielded a 36%-48% relative improvement in AUC. Ablation studies revealed a 2%-22% increase in AUC when compared to simplified architectures lacking key components， substantiating the efficacy of the proposed multimodal fusion strategy and modular design. Incorporation of traditional Chinese medicine （TCM）-specific symptomatology led to an additional 5% improvement in AUC， underscoring the complementary value of integrating TCM and western clinical data. Through gradient-weighted activation mapping visualization analysis， it was found that the model's attention predominantly focused on nodule regions and effectively captured dynamic associations between clinical data and imaging-derived features. ConclusionThe CGN model， by synergistically combining cross-attention encoding with directed graph-based feature integration， enables effective alignment and fusion of heterogeneous multimodal data. The incorporation of both TCM and western clinical information facilitates complementary feature enrichment， thereby enhancing predictive accuracy for pulmonary nodule progression. This approach holds significant potential for supporting intelligent risk stratification and personalized surveillance strategies in lung cancer prevention.

ObjectiveTo construct a comprehensive evaluation indicator system for clinical practice guidelines of traditional Chinese medicine （TCM） that is scientific， systematic， and reflects the characteristics of TCM. MethodsA systematic search was conducted in Chinese and English databases， including CNKI， Wanfang， VIP， SinoMed， PubMed， Embase， and Cochrane Library， to include literature on domestic and international guideline evaluation tools and TCM-related research. Document analysis and CiteSpace were utilized for keyword co-occurrence and clustering analysis. ResultsA total of 65 relevant studies were included， from which seven core thematic domains were identified. Based on the research objectives， a two-step construction strategy was adopted： first， an external evaluation framework was established by referencing international tools to cover methodological rigor and procedural standardization； second， an internal evaluation framework was developed to reflect the distinctive features of TCM clinical practice， including syndrome differentiation and efficacy feedback. Through expert consensus， the indicator system was refined， resulting in a dual-layered structure comprising 8 primary indicators， 22 secondary indicators， and 62 evaluation criteria. ConclusionThe comprehensive evaluation system for TCM clinical practice guidelines， based on bibliometric analysis and core element extraction， integrates both theoretical integrity and practical applicability. This study provides a preliminary research foundation for further optimization， validation， and development of a refined comprehensive evaluation system.

BACKGROUND: Skin cancer is a common skin disease whose incidence and mortality rates have been showing yearly increases. In this report, we update the most recent data on skin cancer as obtained from GLOBOCAN 2022. METHODS: The incidence and mortality rates of skin cancer (melanoma of skin and non-melanoma skin cancer) in GLOBOCAN 2022 were reviewed. These data were analyzed and the characteristics of incidence and mortality across five continents and top five countries and regions in each continent are presented. In addition, correlations between Human Development Index (HDI) and age-standardized incidence and mortality rates of these two skin cancers are described. RESULTS: The GLOBOCAN 2022 data indicated that melanoma was the 17th most common cancer. An estimated 331,722 people were diagnosed with melanoma globally and approximately 58,667 died from this disease. For non-melanoma skin cancer, it ranks as the 5th most common cancer, and an estimated 1,234,533 people were diagnosed with non-melanoma skin cancer globally and approximately 69,416 died from this disease. The incidence of skin cancer varies across geographic regions and countries, with a predominance observed in Oceania, North America, and Europe. Australia was ranked first in terms of incidence, while incidence rates in Africa and Asia were very low. Despite these regional differences in incidence, there was little geographic variation in mortality rates. Currently, the number of deaths from non-melanoma skin cancer exceeds that of melanoma of skin. HDI was positively associated with the incidence of both types of skin cancers, with a positive correlation obtained between HDI and mortality from melanoma of skin and a negative correlation between HDI and mortality from non-melanoma skin cancer. CONCLUSIONS Skin cancer remains a major disease burden worldwide. Substantial variations are observed across countries and regions. Further research on skin cancer will be required to provide a rationale for more effective preventions and treatments of this condition.
Humans ; Skin Neoplasms/mortality* ; Incidence ; Melanoma/mortality* ; Prevalence ; Global Health ; Male ; Female

Objective To explore the association between the relative expression level of SNORD55 in peripheral blood and the outcomes of all-cause mortality and stroke in patients with atrial fibrillation(AF),and to evaluate the predictive value of SNORD55 for prognosis.Methods A total of 133 patients with non-valvular AF admitted in Department of Cardiology of the First Affiliated Hospital of Army Medical University from January 2014 to December 2017 were enrolled in this study.Their baseline information was collected,and the relative expression level of plasma SNORD55 was detected.Cox proportional hazards model was used to explore the association between the relative expression level of SNORD55 in peripheral blood and all-cause mortality as well as stroke in the patients.The predictive performance of CHA2DS2-VASc score for all-cause mortality and stroke was compared with the score combined with the relative expression level of SNORD55 in the AF patients.The area under the receiver operating characteristic curve(AUC)was utilized to evaluate the discrimination,and the net reclassification index(NRI)and comprehensive discriminant improvement index(IDI)were calculated to evaluate the improvement of reclassification ability.Decision curve analysis(DCA)was applied to analyze the change in clinical net benefit.Results The results of multivariate Cox regression showed that high expression of SNORD55 in peripheral blood was an independent risk factor for all-cause mortality and stroke in the AF patients.In predicting the outcomes of all-cause mortality and stroke,the addition of relative expression SNORD55 level with the CHA2DS2-VASc score obtained higher AUC value[0.80(95％CI:0.67～0.93)vs 0.67(95％CI:0.53～0.81),P＜0.05].In predicting the outcome of all-cause death and stroke,combination of the relative expression level of SNORD55 with CHA2DS2-VASc score increased both NRI[54.3(95％CI:10.6～61.9)vs 31.9(95％CI:2.8～47.5),P＜0.05]and IDI[16.1(95％CI:2.4～27.0)vs 7.9(95％CI:0.5～14.8),P＜0.05].The results of DCA showed that our combination of CHA2DS2-VASc score relative expression level of SNORD55 had higher clinical net benefits than the foreign ABC score in the prediction of the outcomes.Conclusion Peripheral blood SNORD55 level is an independent risk factor for all-cause mortality and stroke in AF patients,and has good predictive performance for all-cause mortality and stroke in the patients.

Objectives To explore the association of peripheral blood PIWI-interacting RNA,piR-hsa-2700592,with all-cause mortality and stroke outcomes in patients with atrial fibrillation(AF),and to determine whether piR-hsa-2700592 has the potential to be an AF biomarker.Methods A total of 127 patients with non-valvular AF were enrolled,and the relative expression level of plasma piR-hsa-2700592 was detected.Cox proportional hazard regression was used to analyze the correlation between the expression of piR-hsa-2700592 and all-cause death as well as stroke outcome in the patients.Then the molecule expression level was combined with CHA2DS2-VASc score and ABC stroke(or death)score to establish 2 new prediction models,the improvement of the predictive performance was compared and analyzed.Receiver operating characteristic(ROC)curve analysis(area under the curve,AUC),net reclassification index(NRI),and comprehensive discriminant improvement index(IDI)were used to evaluate the predictive performance,and decision curve analysis(DCA)was employed to assess the clinical benefit.Results Multivariate Cox regression analysis showed that the patients with higher expression level of piR-hsa-2700592 in peripheral blood had a higher risk of stroke(HR:2.203,95%CI:1.120～4.332;P=0.022).In the stroke outcome,combination of plasma piR-hsa-2700592 expression level with CHA2DS2-VASc score and ABC stroke score obtained an AUC of 0.70(95%CI:0.55～0.85,P＜0.001)and 0.84(95%CI:0.73～0.96,P=0.02),respectively.But,no significant association was observed between high plasma piR-hsa-2700592 level and all-cause mortality in the AF patients(HR:1.997;95%CI:0.884～4.509;P=0.096).Combination of plasma piR-hsa-2700592 level improved the discriminative capability than the single CHA2DS2-VASc score and ABC stroke score models,with an NRI and IDI value of 44.20%(95%CI:3.40～59.90,P＜0.001)and 8.20%(95%CI:0.60～15.40,P＜0.001),respectively for the new CHA2DS2-VASc score model,and an NRI and IDI value of 44.20%(95%CI:9.80～58.90,P＜0.001)and 10.40%(95%CI:0.70～21.40,P＜0.001),respectively for the new ABC stroke score model.The DCA curve showed that both new prediction models obtained better net clinical benefits.Conclusion High peripheral blood expression of piR-hsa-2700592 is an independent risk factor for stroke in the AF patients,and the indicator has a good predictive value for prognosis of the patients.piR-hsa-2700592 might be used as a potential biomarker in the diagnosis and prevention of cardiovascular diseases.