Objective To evaluate the performance of lightweight Chinese large language models (LLMs) in answering specialized lung cancer questions, and to explore the impact of retrieval-augmented generation (RAG) on model performance. Methods Eleven lightweight Chinese LLMs with parameter sizes ranging from 7B to 32B were included. A lung cancer-specific evaluation dataset consisting of 200 questions [100 A1-type (basic knowledge) and 100 A2-type (clinical case) questions], constructed based on clinical guidelines and thoracic surgery textbooks, was used for assessment. Model performance was evaluated under two conditions (with and without RAG). Accuracy was used to assess model performance, and response latency was recorded to reflect inference efficiency. An accuracy–latency scatter plot was constructed for descriptive analysis of overall model performance. Results All models successfully completed the evaluation. With the introduction of RAG, the overall average accuracy improved from 61.68% to 76.36%. Smaller models demonstrated the most significant improvement (e.g., the accuracy of DeepSeek-7B increased from 32.50% to 60.00%, P<0.001). The average response latency increased from 12.58 s to 13.80 s. The Qwen3 series showed the best overall performance, and Qwen3-32B achieved the highest accuracy under both conditions (76.50% and 84.00%, respectively). After RAG integration, performance differences among model families were markedly reduced. Based on the accuracy-latency trade-off, Qwen3-32B achieved the best balance between accuracy and response latency under the baseline condition, whereas Qwen3-14B demonstrated superior overall performance in terms of accuracy, latency, and computational cost after RAG integration. Conclusion The integration of RAG technology improves the ability of lightweight Chinese LLMs to answer specialized lung cancer questions. Under the dual practical constraints of limited computational resources and medical data security requirements, the "lightweight model+RAG" technical framework may represent a promising deployment solution.

OBJECTIVES: The major histocompatibility complex class I chain-related gene A (MICA), a component of the human leukocyte antigen (HLA) gene complex, is involved in the pathogenesis of various diseases including cancers and autoimmune disorders. Rosacea, a chronic inflammatory skin disease with a complex pathogenesis, potentially influenced by genetic and autoimmune factors. This study aims to investigate the relationship among MICA gene polymorphisms, single nucleotide polymorphisms (SNPs), and susceptibility to rosacea, thereby offering new insights into the disease mechanism. METHODS: Peripheral blood DNA samples were collected from 84 patients with rosacea (rosacea group) and 223 healthy volunteers (control group) who visited the Dermatology Outpatient Department of Xiangya Hospital of Central South University between November 2017 and November 2019. MICA genotyping was performed using polymerase chain reaction-sequencing-based typing (PCR-SBT) and the next-generation sequencing (NGS), and the accuracy of the 2 methods was compared. The frequency distributions of MICA alleles between the 2 groups were analyzed. Amino acid clustering and SNP site analyses were conducted to identify haplotype-linked SNPs and to classify MICA polymorphic variants. Distribution differences of these classifications between groups were also examined. RESULTS: Blood tests in rosacea patients showed mildly elevated, with no significant changes in lymphocyte counts. Both PCR-SBT and NGS accurately identified MICA alleles. The most common alleles in the rosacea group were MICA*010:01, MICA*008:04, and MICA*019:01. The frequencies of MICA*002:01 and MICA*027 were significantly lower in the rosacea group compared to controls (6.55% vs 18.16% and 1.19% vs 5.38%, respectively), while and MICA*010:01 were significantly higher (7.74% vs 3.36% and 31.55% vs 18.61%, respectively; all P<0.05). Five short tandem repeat (STR) alleles were identified. Frequencies of MICA-A4 and MICA-A9 were lower in the rosacea group than in the control group (16.07% vs 23.32% and 7.74% vs 17.26%, respectively), whereas MICA-A6 was higher (10.12% vs 4.03%; all P<0.05). Clustering and SNP analysis identified 6 linked SNP sites, classifying MICA variants into Type I (C₃₆+M₁₂₉+K₁₇₃+G₂₀₆+W₂₁₀+S₂₁₅) and Type II (Y₃₆+V₁₂₉+E₁₇₃+S₂₀₆+R₂₁₀+T₂₁₅). Type I MICA variants were significantly associated with rosacea susceptibility. CONCLUSIONS MICA gene polymorphisms are associated with susceptibility to rosacea, and there are 6 linked SNP sites within the MICA gene. Based on this, MICA polymorphic variants are classified into Type I and Type II, with Type I being more closely associated with disease development of rosacea.
Humans ; Polymorphism, Single Nucleotide ; Histocompatibility Antigens Class I/genetics* ; Rosacea/genetics* ; Genetic Predisposition to Disease/genetics* ; Female ; Male ; Adult ; Middle Aged ; Genotype ; Alleles ; Gene Frequency ; Haplotypes ; Case-Control Studies ; Aged ; High-Throughput Nucleotide Sequencing

OBJECTIVES: To construct a nomogram for predicting the efficacy of immune checkpoint inhibitors (ICIs) in advanced non-small cell lung cancer (aNSCLC) by integrating chest CT radiomics signature that reflects the tumor microenvironment (TME) and clinical parameters of the patients. METHODS: Transcriptomic and CT imaging data from TCGA, GEO and TCIA databases were integrated for weighted gene co-expression network analysis (WGCNA) of the GEO cohort to identify the immunotherapy-related genes (IRGs) associated with ICIs response. A prognostic model was built using these IRGs in the TCGA cohort to assess immune microenvironment features across different risk groups. Radiomics features were extracted from TCIA lung_3 cohort using PyRadiomics, and 94 features showing strong association with IRGs (|r|>0.4) were selected. A retrospective cohort consisting of 210 aNSCLC patients receiving first-line ICIs at Guangdong Provincial People's Hospital was analyzed and divided into training (n=147) and validation (n=63) groups. Least absolute shrinkage and selection operator was used for radiomic features selection, and logistic regression was applied to construct a combined clinical-radiomic model and nomogram for predicting ICIs therapy response. The performance of the model was evaluated using ROC curve, calibration curve, and decision curve analysis. RESULTS: WGCNA identified 84 IRGs enriched in immune activation pathways. The combined model outperformed individual models in both the training (AUC=0.725, 95% CI: 0.644-0.807) and validation cohorts (AUC=0.706, 95% CI: 0.577-0.836). Calibration curve and decision curve analyses confirmed the clinical efficacy of the nomogram for predicting ICIs therapy response in aNSCLC patients. CONCLUSIONS The genomic-radiomic-clinical multidimensional predictive framework established in this study provides an interpretable biomarker combination and clinical decision-making tool for evaluating ICIs efficacy in aNSCLC, potentially facilitating personalized immunotherapy decision-making.
Humans ; Carcinoma, Non-Small-Cell Lung/therapy* ; Tumor Microenvironment ; Lung Neoplasms/therapy* ; Immunotherapy ; Tomography, X-Ray Computed ; Nomograms ; Retrospective Studies ; Immune Checkpoint Inhibitors/therapeutic use* ; Prognosis ; Male ; Female ; Radiomics

Objective To analyze the incidence,colonization pathways,and predictive factors of bacterial colonization of epidural analgesia catheters in patients with chronic pain.Methods A total of 150 patients with chronic pain who underwent continuous epidural catheterization(catheter in-dwelling time of 7 to 10 days)were selected as study subjects.Samples from three sites were collect-ed for bacterial culture.Clinical data of the patients were collected,and the positive rate of bacterial culture,characteristics of bacterial species distribution,and bacterial colonization pathways were ana-lyzed.The efficacy of predictive factors was assessed using the receiver operating characteristic(ROC)curve.Results The positive rates of bacterial culture in samples from the skin swabbing fluid around the puncture site,the subcutaneous segment of the catheter,and the catheter tip were 22.0％,7.3％,and 8.7％,respectively.Staphylococcus epidermidis was the predominant colonizing bacterial species.Spearman correlation coefficient analysis showed a significant correlation between the results of bacterial culture from the skin around the puncture site and catheter tip colonization(r=0.47,P＜0.01).ROC curve analysis revealed that the area under the curve of bacterial culture results from the skin around the puncture site in predicting catheter tip bacterial colonization was 0.843,with a sensitivity of 84.9％and a specificity of 84.6％.Conclusion Bacterial migra-tion along the catheter is the main pathway for catheter tip bacterial colonization,and the results of bacterial culture from the skin around the puncture site are an effective predictive factor for the risk of bacterial colonization.

Objective To observe the value of shear wave elastography(SWE)for evaluating therapeutic effect of ultrasound-guided drug injection for muscle injury.Methods Eighty patients with unilateral muscle injury were retrospectively included,including 40 cases underwent ultrasound-guided drug injection(group A)and 40 cases underwent electromagnetic wave physiotherapy plus external application of Yunnan Baiyao Gao(group B).Pain intensity was assessed using visual analogue scale(VAS)before treatment and 3 weeks after the final treatment,while the Young modulus(E)value of the injured muscle was measured before treatment and 1,2 and 3 weeks after final treatment,and the improvements of VAS scores and E values were compared between groups.Results The total effective rate in group A(35/40,87.50%)was higher than that in group B(21/40,52.50%;P＜0.05).Before treatment,no significant difference of VAS score was found between group A(8.07±0.83)and group B(7.88±0.85)(P＞0.05).After treatment,VAS scores decreased in both groups(both P＜0.05),which in group A(2.30±1.07)was more obviously than that in group B(4.80±0.82)(P＜0.05).After treatment,E values of injury muscles increased significantly in both groups(P＜0.05),while group A had a greater increase in overall magnitude and overall rate than group B(P＜0.05).Conclusion Ultrasound-guided drug injection therapy had significant therapeutic effect for muscle injuries,which could be dynamically monitored with SWE.

Objective To explore the feasibility and safety of fastest recovery after surgery(FRAS)in laparoscopic surgery of gastrointestinal tumors.Methods The clinical data of patients undergoing laparoscopic surgery for gastrointestinal tumors under FRAS and enhanced recovery after surgery(ERAS)from Jan.2023 to May 2024 were collected,and perioperative safety and medical cost were analyzed.Results A total of 87 patients were enrolled,including 43 in the FRAS group and 44 in the ERAS group.Compared with the ERAS group,the FRAS group had significantly shorter surgical time(3.0[2.5,3.5]h vs 3.0[2.5,4.0]h),first postoperative movement time([2.85±4.29]h vs[20.18±6.13]h),first postoperative oral feeding time(2.0[2.0,3.0]h vs 24.0[15.0,48.0]h),postoperative hospital stay(24.0[20.0,40.0]h vs 192.0[150.0,216.0]h),lower hospitalization costs(50 515.61[46 650.44,56 827.12]yuan vs 65 555.09[58 683.21,86 239.02]yuan),and lower medication costs(2 671.09[2 063.31,3 127.09]yuan vs 7 326.90[5 104.66,10 674.26]yuan)(all P＜0.05).Conclusion It is safe and feasible to use FRAS during the perioperative period of laparoscopic radical gastrectomy for gastrointestinal tumors,and FRAS can also reduce the costs of hospitalization and medications.

Objective To investigae microbial community structure of lower respiratory tract in severe hospital-acquired pneumonia(SHAP)patients in intensive care unit(ICU)using metagenomic next-generation sequencing(mNGS).Methods mNGS was performed on bronchoalveolar lavage fluid(BALF)of 84 patients with SHAP.Patients were grouped based on age,smoking status,underlying diseases and duration of artificial airway.The differences in α diversity,β diversity,microbial composition and community structure of airway microbiota were compared between different groups.The differential airway microbiota associated with artificial airway were screened,and microbial co-occurrence networks was constructed to observe the interaction in microorganisms.Results Results of α diversity analysis revealed that diversity and evenness of the microbial community were higher in young adults compared to those of middle-aged patients,while microbial diversity and evenness were significantly reduced in patients with comorbid stroke.In the group of artificial airway treatment,the diversity and uniformity of microorganisms decreased as the duration of artificial airway treatment increased.The diversity was the lowest when the artificial airway treatment lasted for more than 3 days.β diversity analysis confirmed that there were significant differences in the distinct microbial community structures between the>3 days support group and the non-intubated and≤3 days support cohorts.Acinetobacter baumannii was dominant in all groups.The bacterial diversity was significantly higher in the middle-aged group,the non-smoking group,the group without artificial airway therapy,the group without diabetes mellitus,the group with artificial airway therapy≥3 days and the group with chronic lung disease than those of other groups.In particular,pseudomonas aeruginosa,corynebacterium striatum and veillonella parvula were enriched in these groups.Difference analysis showed that there were significant differences in pseudomonas aeruginosa and corynebacterium striatum between the group with artificial airway treatment>3 days,the group without artificial airway therapy and the group with artificial airway treatment≤3 days.Network co-occurrence showed that there may be synergistic or antagonistic relationships between some microorganisms.Conclusion The microbial diversity of the lower respiratory tract in patients with SHAP significantly decreases in the elderly,those with concurrent stroke and those receiving artificial airway treatment.For these groups,rational use of antibiotics should be adopted to guide precise anti-infection treatment.

Objective To observe the value of shear wave elastography(SWE)for evaluating therapeutic effect of ultrasound-guided drug injection for muscle injury.Methods Eighty patients with unilateral muscle injury were retrospectively included,including 40 cases underwent ultrasound-guided drug injection(group A)and 40 cases underwent electromagnetic wave physiotherapy plus external application of Yunnan Baiyao Gao(group B).Pain intensity was assessed using visual analogue scale(VAS)before treatment and 3 weeks after the final treatment,while the Young modulus(E)value of the injured muscle was measured before treatment and 1,2 and 3 weeks after final treatment,and the improvements of VAS scores and E values were compared between groups.Results The total effective rate in group A(35/40,87.50%)was higher than that in group B(21/40,52.50%;P＜0.05).Before treatment,no significant difference of VAS score was found between group A(8.07±0.83)and group B(7.88±0.85)(P＞0.05).After treatment,VAS scores decreased in both groups(both P＜0.05),which in group A(2.30±1.07)was more obviously than that in group B(4.80±0.82)(P＜0.05).After treatment,E values of injury muscles increased significantly in both groups(P＜0.05),while group A had a greater increase in overall magnitude and overall rate than group B(P＜0.05).Conclusion Ultrasound-guided drug injection therapy had significant therapeutic effect for muscle injuries,which could be dynamically monitored with SWE.

Objective To investigae microbial community structure of lower respiratory tract in severe hospital-acquired pneumonia(SHAP)patients in intensive care unit(ICU)using metagenomic next-generation sequencing(mNGS).Methods mNGS was performed on bronchoalveolar lavage fluid(BALF)of 84 patients with SHAP.Patients were grouped based on age,smoking status,underlying diseases and duration of artificial airway.The differences in α diversity,β diversity,microbial composition and community structure of airway microbiota were compared between different groups.The differential airway microbiota associated with artificial airway were screened,and microbial co-occurrence networks was constructed to observe the interaction in microorganisms.Results Results of α diversity analysis revealed that diversity and evenness of the microbial community were higher in young adults compared to those of middle-aged patients,while microbial diversity and evenness were significantly reduced in patients with comorbid stroke.In the group of artificial airway treatment,the diversity and uniformity of microorganisms decreased as the duration of artificial airway treatment increased.The diversity was the lowest when the artificial airway treatment lasted for more than 3 days.β diversity analysis confirmed that there were significant differences in the distinct microbial community structures between the>3 days support group and the non-intubated and≤3 days support cohorts.Acinetobacter baumannii was dominant in all groups.The bacterial diversity was significantly higher in the middle-aged group,the non-smoking group,the group without artificial airway therapy,the group without diabetes mellitus,the group with artificial airway therapy≥3 days and the group with chronic lung disease than those of other groups.In particular,pseudomonas aeruginosa,corynebacterium striatum and veillonella parvula were enriched in these groups.Difference analysis showed that there were significant differences in pseudomonas aeruginosa and corynebacterium striatum between the group with artificial airway treatment>3 days,the group without artificial airway therapy and the group with artificial airway treatment≤3 days.Network co-occurrence showed that there may be synergistic or antagonistic relationships between some microorganisms.Conclusion The microbial diversity of the lower respiratory tract in patients with SHAP significantly decreases in the elderly,those with concurrent stroke and those receiving artificial airway treatment.For these groups,rational use of antibiotics should be adopted to guide precise anti-infection treatment.

Objective: To examine the mediating effect of positive coping style on health literacy and lifestyle among medical students, so as to provide the reference for health education and health promotion among medical students in universities. Methods: Medical students of six full-time medical colleges in Guizhou Province were sampled using a multi-stage stratified cluster random sampling method from April to June 2023, and their gender, age and educational level were investigated by questionnaires. The College Students' Healthy Lifestyle Scale, the Health Literacy Questionnaire and the Simplified Coping Style Scale were used to evaluate lifestyle, health literacy and coping style of students. The Process program was used to analyze the mediating effect of positive coping style on health literacy and lifestyle, and Bootstrap method was used to examine the significance of the mediating effect. Results: A total of 9 266 peoples were investigated, including 3 177 males (34.29%) and 6 089 females (65.71%), and had a mean age of (21.06±2.61) years. There were 2 775 junior college students (29.95%), 4 864 undergraduates (52.49%) and 1 627 master students (17.56%). The scores of lifestyle, health literacy and positive coping style among medical students were 118.34±19.06, 54.57±10.81 and 2.81±0.68, respectively. Health literacy had a direct positive effect on lifestyle (effect value=0.286, 95%CI: 0.257-0.316) and a indirect positive effect on lifestyle through positive coping style (effect value=0.215, 95%CI: 0.193-0.236), and the mediating effect contributed 42.91% to the total effect. Conclusion Health literacy could improve lifestyle indirectly through positive coping style among medical students.