Objective: To develop QingNangTCM, a specialized large language model (LLM) tailored for expert-level traditional Chinese medicine (TCM) question-answering and clinical reasoning, addressing the scarcity of domain-specific corpora and specialized alignment. Methods: We constructed QnTCM_Dataset, a corpus of 100 000 entries, by integrating data from ShenNong_TCM_Dataset and SymMap v2.0, and synthesizing additional samples via retrieval-augmented generation (RAG) and persona-driven generation. The dataset comprehensively covers diagnostic inquiries, prescriptions, and herbal knowledge. Utilizing P-Tuning v2, we fine-tuned the GLM-4-9B-Chat backbone to develop QingNangTCM. A multi-dimensional evaluation framework, assessing accuracy, coverage, consistency, safety, professionalism, and fluency, was established using metrics such as bilingual evaluation understudy (BLEU), recall-oriented understudy for gisting evaluation (ROUGE), metric for evaluation of translation with explicit ordering (METEOR), and LLM-as-a-Judge with expert review. Qualitative analysis was conducted across four simulated clinical scenarios: symptom analysis, disease treatment, herb inquiry, and failure cases. Baseline models included GLM-4-9B-Chat, DeepSeek-V2, HuatuoGPT-II (7B), and GLM-4-9B-Chat (freeze-tuning). Results: QingNangTCM achieved the highest scores in BLEU-1/2/3/4 (0.425/0.298/0.137/0.064), ROUGE-1/2 (0.368/0.157), and METEOR (0.218), demonstrating a balanced and superior normalized performance profile of 0.900 across the dimensions of accuracy, coverage, and consistency. Although its ROUGE-L score (0.299) was lower than that of HuatuoGPT-II (7B) (0.351), it significantly outperformed domain-specific models in expert-validated win rates for professionalism (86%) and safety (73%). Qualitative analysis confirmed that the model strictly adheres to the “symptom-syndrome-pathogenesis-treatment” reasoning chain, though occasional misclassifications and hallucinations persisted when dealing with rare medicinal materials and uncommon syndromes. Conclusion Combining domain-specific corpus construction with parameter-efficient prompt tuning enhances the reasoning behavior and domain adaptation of LLMs for TCM-related tasks. This work provides a technical framework for the digital organization and intelligent utilization of TCM knowledge, with potential value for supporting diagnostic reasoning and medical education.

Objective On the basis of the theory of the tumor metastatic state, this work aims to explore the metastasis trajectories and peripheral immune response patterns in patients with non-small cell lung cancer (NSCLC) receiving vital qi-strengthening and toxin-eliminating therapy (Fuzheng Juandu) and reveal the value of peripheral immunity in assessing vital qi deficiency during metastasis. Methods A retrospective study design was adopted. Patients with NSCLC admitted from January 1, 2018, to October 31, 2025 were enrolled. Data on metastasis status and peripheral immune indices were collected at the follow-up time points of one, two, three, and five years. Group-based trajectory modeling was used to fit metastasis trajectories, and the independent predictors of trajectory grouping were analyzed through multifactor and multiclassification logistic regression. Results A total of 184 patients with NSCLC were included, and three different groups of metastasis trajectories were identified, namely, the low-risk stable (46.74%, n=86), low-risk rising (13.59%, n=25), and high-risk stable (39.67%, n=73) groups. Univariate analysis revealed substantial differences in peripheral immune indices among the three trajectory groups. Multivariate analysis showed that TNM stage, absolute neutrophil count, lymphocyte count, absolute monocyte count, platelet volume, SAA:CRP ratio, and NLR were independent predictors of the low-risk rising group, whereas TNM stage, absolute monocyte count, basophil count, SII, and NLR were independent predictors of the high-risk stable group. Conclusion Under Fuzheng Juandu therapy, patients with NSCLC exhibit three distinct metastasis trajectories, each corresponding to a specific peripheral immune response pattern. Elucidating these peripheral immune response patterns may facilitate the objective quantification of the degree of vital qi deficiency, prediction of future metastasis risk, and providing an important reference for the individualized clinical implementation of Fuzheng Juandu therapy.

OBJECTIVE: To compare the effectiveness of posterior axillary edge approach and arthroscopic assisted reduction in the treatment of Ideberg type Ⅰ and Ⅱ glenoid fracture of the scapula. METHODS: The clinical data of 26 patients with fresh Ideberg type Ⅰ and Ⅱ scapular fractures admitted between June 2021 and September 2024 who met the selection criteria were analyzed retrospectively. The patients were divided into two groups according to different treatment methods. Ten cases in the posterior axillary edge group were fixed by open reduction plate through the posterior axillary edge approach, and 16 cases in the arthroscopy group were treated with suture anchor fixation under arthroscopy. There was no significant difference in baseline data between the two groups ( P>0.05), such as gender, age, surgical side, Ideberg type, cause of injury, time from injury to operation, rotator cuff injury, and superior labrum anterior posterior (SLAP) injury, etc. The operation time and fracture healing time were recorded and compared between the two groups, and the shoulder pain was evaluated by visual analogue scale (VAS) score at 1 week, 1 month, and 3 months after operation. At 3 and 6 months after operation, the range of motion of shoulder joint in anteflexion, abduction, external rotation, internal rotation, and backward extension was evaluated, the upper limb dysfunction was evaluated by the Disability Assessment Scale of Arm, Shoulder, and Hand (DASH), and the shoulder joint function was evaluated by the Constant-Murley score. The differences between 6 months and 3 months after operation (changes) were statistically analyzed. RESULTS: Patients in both groups were followed up 11-13 months, with an average of 12.5 months. The operation time and fracture healing time in the posterior axillary edge group were significantly shorter than those in the arthroscopy group ( P<0.05). There was no complication such as wound infection, vascular and nerve injury, loss of reduction, bone nonunion, or glenohumeral instability in both groups. At 1 week after operation, the VAS score in the posterior axillary edge group was significantly higher than that in the arthroscopy group ( P<0.05); there was no significant difference in the VAS score between the two groups at 1 and 3 months after operation ( P>0.05). At 6 months after operation, the changes of shoulder joint in anteflexion, internal rotation range of motion and DASH scores in the posterior axillary edge group were significantly lower than those in the arthroscopy group ( P<0.05), while the changes of abduction, external rotation, backward extension range of motion and Constant-Murley scores were not significantly different between the two groups ( P>0.05). CONCLUSION For Ideberg type Ⅰ and Ⅱ glenoid fracture of the scapula, the posterior axillary edge approach for internal fixation has a short operation time, fast fracture healing, and is beneficial for early functional recovery; arthroscopic assisted reduction has minimal trauma and can handle joint injuries simultaneously. Both surgical procedures are safe and effective, and individualized selection should be made based on soft tissue conditions and combined injuries.
Humans ; Arthroscopy/methods* ; Scapula/surgery* ; Male ; Female ; Retrospective Studies ; Adult ; Fracture Fixation, Internal/instrumentation* ; Fractures, Bone/surgery* ; Middle Aged ; Treatment Outcome ; Bone Plates ; Suture Anchors ; Fracture Healing ; Range of Motion, Articular ; Young Adult ; Shoulder Joint/surgery* ; Operative Time

ObjectiveTo explore the pathological mechanism-syndrome-treatment patterns of approved Chinese patent medicines （CPMs） that treat collateral disorders， providing a reference for the principle of "treating different diseases with the same therapy" in collateral pathology. MethodsCPMs that apply treatment strategies based on collateral disorders were identified from the Pharmacodia database by extracting information from the "efficacy" or "indications" sections of drug package inserts. A database was established to extract the names and compositions of the CPMs， as well as their indications， related traditional Chinese medicine （TCM） symptoms， disease locations （affected areas）， and pathological factors. Frequency statistics were performed. Using the Apriori algorithm， an association rule analysis was conducted on CPMs and disease-location combinations related to the top three most frequent pathological factor combinations. Core formulas for these combinations were identified and analyzed through drug network analysis and MCODE module clustering. ResultsA total of 660 CPMs targeting collateral disorders were retrieved， involving 299 indications， 323 TCM symptoms， 21 disease locations， 19 pathological factors， and 124 pathological factor combinations. The most frequent pathological factor combinations were blood stasis （involved in 109 CPMs， 16.52%）， exogenous wind （外风） -cold-dampness （involved in 43 CPMs， 6.52%）， and qi deficiency-blood stasis （involved in 42 CPMs， 6.36%）. Analysis of the core formulas for these combinations revealed common ingredients such as Honghua （Carthami Flos）， Chuanxiong （Chuanxiong Rhizoma）， Danggui （Angelicae Sinensis Radix）， and Dilong （Pheretima）. ConclusionCollateral disorders involve a wide range of pathogenesis and represent a fundamental mechanism in the onset and development of various diseases， characterized by obstruction and stagnation. The primary therapeutic principle is unblocking of the collaterals. Blood stasis obstructing the collaterals is the core pathological basis， and the strategy of activating blood circulation and resolving stasis to unblock the collaterals should be central to the treatment. The core medication pattern involves combining blood-activating and stasis-resolving herbs with insect-derived medicinals that unblock collaterals. Exogenous wind is often the initiating patholo-gical factor in colla-teral disorders， and the appropriate addition of wind-dispelling herbs can enrich the treatment strategies for such conditions.

IgA nephropathy (IgAN) is the most common primary glomerulonephritis. Mesangial hyperplasia and deposition of IgA immune complex are typical pathological changes of IgAN. Animal model is an important tool to explore the pathogenesis, diagnosis and treatment plan, and evaluate the safety and efficacy of drugs. At present, there are many kinds of IgAN animal models, including humanized animal models and non-humanized animal models, and there are great differences in the modeling principle, method, time and pathological changes. The humanized animal model is closer to the pathogenesis of IgAN in humans and has gradually become a research hotspot. In this paper, common animal models of IgAN in detail from the aspects of modeling method and time, characteristics and significance of each model, and the research progress of anthropomorphic animal models are reviewed to provide reference and inspiration for the basic research of IgAN.

Objective To explore the targets and mechanism of Buyang Huanwu Decoction in preventing and treating atherosclerosis through network pharmacology,molecular docking and animal experiment.Methods The main active components and related targets of Buyang Huanwu Decoction were obtained and screened through databases such as TCMSP,HERB,UniProt,PubChem,SwissADME and SwissTargetPrediction.Six disease databases such as DrugBank,GeneCards,DisGeNET,OMIM,PharmGKB and TTD were used to obtain AS related targets.R language was used to obtain common targets for the drug and disease.STRING 11.5 database was used for protein interaction network analysis,and Cytoscape 3.9.1 software was used to conduct network topology parameters and obtain core targets.GO and KEGG enrichment analysis were performed on the common targets using DAVID platform.Molecular docking of core genes with the main active components of Buyang Huanwu Decoction was conducted using AutoDock Vina 1.5.6 software.An AS mouse model was established and intervened with Buyang Huanwu Decoction.The lipid content of the mouse aortic sinus was observed using Oil Red O staining.Western blot was used to detect the expression of PI3K/AKT/p38 signaling pathway,and ELISA was used to detect the contents of TNF-α and IL-17 in mouse serum.Results A total of 104 main active components and 283 action targets of Buyang Huanwu Decoction were obtained,as well as 5 347 AS related targets and 218 common targets for drugs and disease.Buyang Huanwu Decoction may use key active components such as quercetin,kaempferol and baicalin as core targets for TP53,MAPK1,AKT1,and exert its therapeutic effect on AS through PI3K-Akt signaling pathway,TNF signaling pathway and IL-17 signaling pathway,etc.Molecular docking indicated that quercetin,luteolin,kaempferol and baicalin had good affinity with TP53,HSP90AA1,MAPK1,AKT1 and RELA targets.MAPK1 had strong binding activity with baicalin and luteolin.The experimental results showed that Buyang Huanwu Decoction could reduce the lipid content in aortic sinus of AS mice,reduce the contents of TNF-α and IL-17 in serum(P<0.01)and significantly down-regulate the expression levels of PI3K,p-AKT and p-p38 proteins(P<0.05,P<0.01).Conclusion The mechanism of Buyang Huanwu Decoction for preventing and treating AS may be related to regulating the expression of the PI3K/AKT/p38 signaling pathway,thereby reducing inflammatory response.

The establishment of mouse models is critical for discovering the biological targets of tumorigenesis and cancer development, preclinical trials of targeted drugs, and formulation of personalized therapeutic regimens. Currently, the patient-derived xenograft (PDX) model is considered a reliable animal tumor model because of its ability to retain the characteristics of the primary tumor at the histopathological, molecular, and genetic levels, and to preserve the tumor microenvironment. The application of the PDX model has promoted in-depth research on tumors in recent years, focusing on drug development, tumor target discovery, and precise treatment of patients. However, there are still some common questions. This review introduces the latest research progress and common questions regarding tumors with high mortality rates, focusing on their application in targeted drug screening and the formulation of personalized medical strategies. The challenges faced, improvement methods, and future development of the PDX model in tumor treatment applications are also discussed. This article provides technical guidance and comprehensive expectations for anti-cancer drug screening and clinical personalized therapy.

Dental treatments for children and adolescents have unique clinical characteristics that differ from dental care for adults in terms of children's physiology, psychology, and behavior. These differences impose specific requirements on the application of local anesthesia in pediatric dental procedures. This article presents expert consensus on the principles of local anesthesia techniques in pediatric dental therapies, including the use of common anesthetic drugs and dosage control, safety and efficacy evaluation, and prevention and management of complications. The aim is to improve the safety and quality of pediatric dental treatments and offer guidance for clinical application by dentists.
Humans ; Child ; Anesthesia, Local/methods* ; Consensus ; Anesthesia, Dental/methods* ; Adolescent ; Anesthetics, Local/administration & dosage* ; Dental Care for Children

Objective:To construct and evaluate the predictive performance of a multiphase ultrasound radiomics model for microvascular invasion（MVI）in hepatocellular carcinoma（HCC）.Methods:A total of 126 patients with pathologically confirmed HCC were retrospectively enrolled from 4 medical centers between May 2018 and July 2025，including the First Hospital of Shanxi Medical University，Shanxi Province Third People's Hospital，Changzhi People's Hospital，and the Organ Transplant Center of the Affiliated Hospital of Qingdao University. A total of 630 ultrasound images of the lesions in different phases were collected，from which 1 561 radiomic features were extracted. The patients from medical institutions in Shanxi Province were chosen as the training set（ n=91），and the patients from the Organ Transplant Center of the Affiliated Hospital of Qingdao University were chosen as the validation set（ n=35）. In the training set，37.4%（34/91）patients presented MVI（+），whereas in the validation set，54.3%（19/35）patients presented MVI（+）. Radiomics features were extracted from ultrasound images，and features related to the MVI（+）were selected through dimensionality reduction analysis. Five multiple machine learning algorithms were used to construct predictive models，which were then evaluated using an external validation set. The Radscore was calculated，and a nomogram was constructed combining Radscore with ultrasound and clinical characteristics to predict MVI. Results:The model combining radiomics features from the portal venous phase and the delay phase showed the best predictive performance in both the training and validation sets，with area under curve（AUC）values of 0.835 and 0.727，respectively. The prediction model developed using radiomics Radscore and clinical indicators could be represented and presented as a nomogram.Conclusions:The radiomics model based on multi-phase ultrasound offers a novel approach for non-invasive preoperative prediction of MVI in liver cancer. Furthermore，its integration with clinical features aids in optimizing clinical treatment strategies.

Objective:To develop and evaluate a medical visual question answering (VQA) model for neonatal lung ultrasound (LUS) images to enhance intelligent auxiliary diagnosis of neonatal pulmonary diseases.Methods:Using data from neonates admitted to Beijing Obstetrics and Gynecology Hospital, Capital Medical University (January 2023 to December 2024), an image-question-answer dataset comprising 251 LUS images was constructed [43 pneumonia (17.1%), 42 neonatal respiratory distress syndrome (16.7%), 83 transient tachypnea (33.1%), and 83 normal (33.1%) images] with a four-tier medical question-answer framework. Building upon the Qwen2.5-VL-7B base model and integrating LoRA fine-tuning with chain-of-thought prompting, we developed the NLUS-VQA model to enhance visual-language semantic alignment and enable stepwise clinical reasoning, achieving efficient small-sample adaptation. Model performance was comprehensively assessed through natural language generation metrics (BLEU-4, ROUGE-1/2/L), qualitative evaluation of characteristic recognition, and clinical consistency analysis.Results:(1) Quantitative evaluation demonstrated that NLUS-VQA achieved scores of 22.38 (BLEU-4), 48.26 (ROUGE-1), 22.40 (ROUGE-2), and 37.20 (ROUGE-L), representing significant improvements over baseline models. (2) Qualitatively, the model exhibited strong performance in identifying lung consolidation, coalescent B-lines, and snowflake signs, with its chain-of-thought strategy enhancing clinical interpretability and answer accuracy. (3) Clinically, NLUS-VQA achieved a Cohen's Kappa coefficient of 0.78 and diagnostic accuracy of 80.8% (21/26), indicating substantial agreement with clinical experts.Conclusion:The NLUS-VQA model demonstrates robust interpretability in recognizing key sonographic patterns (e.g. lung consolidation, confluent B-lines, and snowflake signs), providing a scalable framework for small-sample medical image analysis, though diagnostic performance on complex conditions remains limited by dataset scale and minority class representation.