1.Construction and practice of application model for localized large language model in preoperative medication reconciliation for gastric cancer
Yuxuan ZHU ; Jizhong ZHANG ; Yuhao SUN ; Jiayu WEN ; Xin LIU ; Jifu WEI ; Lingli HUANG
China Pharmacy 2026;37(8):1062-1067
OBJECTIVE To construct a preoperative medication reconciliation model assisted by a localized large language model (LLM) for gastric cancer and evaluate its clinical efficacy. METHODS A total of 249 gastric cancer patients with a history of continuous medication before admission in the Gastric Surgery Department of Jiangsu Cancer Hospital were retrospectively enrolled. Patients were divided into training set (154 cases) and validation set (95 cases) based on the order of time. Based on guidelines, drug package inserts, and other evidence, a standardized medication reconcili ation process and a structured knowledge base were constructed. DeepSeek-V3 LLM was deployed privately in the hospital, combined with retrieval-augmented generation technology, to achieve automated integration of medication information, risk screening, and generation of personalized recommendations. The quality of LLM-generated recommendations was evaluated using automatic metrics (BERT Score and ROUGE-1, 2, L) and manual scoring [seven-dimensional index (7DI) ] . Spearman correlation analysis was performed to explore the correlation between automatic scores and manual scores. Cronbach’s α coefficient was used to test the internal consistency of manual scoring results. The time consumed by manual and LLM-assisted medication reconciliation was compared across tasks of different difficulty levels (simple, moderate, and high). RESULTS A structured knowledge base covering 8 major drug categories was finally established, covering common and high-risk preoperative medication scenarios and providing structured retrieval support for the LLM. For automatic evaluation, the precision, recall, and F1-score of BERT Score were 0.783±0.033, 0.811±0.038, and 0.796±0.028, respectively. The F1-scores of ROUGE-1, ROUGE-2 and ROUGE-L were 0.566±0.067, 0.338±0.076 and 0.468±0.082, respectively. The 7DI scores from three manual raters ranged from 32.06 to 33.45. The F1-score of automatic scoring was significantly positively correlated with the 7DI score of manual scoring (maximum coefficient of determination=0.611, P <0.001), and the internal consistency of manual scoring was good (Cronbach’s α = 0.876). In terms of efficiency, LLM-assisted medication reconciliation reduced time consumption by more than 90% compared with manual reconciliation in the simple, moderate, and high-difficulty groups ( P <0.001). CONCLUSIONS The medication reconciliation model constructed based on a localized LLM and structured knowledge base shows high accuracy, consistency, and clinical applicability in complex preoperative medication scenarios for gastric cancer. It can improve the efficiency of medication reconciliation and reduce potential medication risks.
2.Protective effect of the active component DMDD from Averrhoa carambola root on myocardial injury in diabetic mice and its correlation with the NCOA4/FTH1/ATG8 axis
Yongxin CHEN ; Yuxuan LI ; Kailei GU ; Jiajun YOU ; Xiaohan SUN ; Jing MA ; Yanping ZHOU ; Xiaojie WEI
China Pharmacy 2026;37(9):1141-1147
OBJECTIVE To investigate the protective effect of 2-dodecyl-6-methoxy-2,5-diene-1,4-cyclohexanedione (DMDD), an active component from Averrhoa carambola root, on myocardial injury in diabetic mice based on the nuclear receptor coactivator 4/ferritin heavy chain 1/autophagy-related protein 8 (NCOA4/FTH1/ATG8) axis. METHODS The successfully modeled diabetic mice were randomly divided into model group and DMDD low-, medium-, and high-dose (12.5, 25, 50 mg/kg) groups, while an additional non-modeled control group was established, with 6 mice in each group. Each group received the corresponding drug solution or an equal volume of normal saline intragastically once daily for 21 consecutive days. After the administration, the levels of fasting blood glucose (FBG), serum lactate dehydrogenase (LDH), and creatine kinase isoenzyme MB (CK-MB) were measured. Myocardial pathological changes, degree of fibrosis, and myocardial cell ultrastructure were observed. Myocardial cell death index and NCOA4 protein positive index were detected. The protein expression levels of NCOA4, FTH1, ATG8, solute carrier family 7 member 11 (SLC7A11), and glutathione peroxidase 4 (GPX4) in cardiac tissue were measured. RESULTS Compared with model group, each DMDD group showed significant alleviation of cardiac pathological injury and varying degrees of improvement in the myocardial cell ultrastructure. The FBG and serum LDH and CK-MB levels, the myocardial cell death index and NCOA4 protein positive index,the protein expression levels of NCOA4, FTH1, and ATG8 in cardiac tissue were significantly decreased ( P <0.001), while the protein expression levels of SLC7A11 and GPX4 were significantly increased ( P <0.001). CONCLUSIONS DMDD can reduce blood glucose levels, alleviate myocardial histopathological injury, and inhibit cell death in diabetic mice. The mechanism is associated with inhibiting excessive activation of the NCOA4/FTH1/ATG8 axis and reducing ferritinophagy.
3.Research Progress of 223-Ra in the Treatment of Bone Metastases from Desmoplasia-resistant Prostate Cancer
Chang LU ; Ran ZHANG ; Li ZHANG ; Jiaxin DING ; Yue SUN ; Zhuoling RAN ; Yuxuan ZHENG ; Lin YU ; Xu GAO ; Jing XIE ; Huan ZHOU ; Jian GONG
Herald of Medicine 2025;44(3):446-451
Prostate cancer is one of the most common male urological malignancies,in which bone metastasis of desmo-plasia-resistant prostate cancer is an important stage in the progression of the disease,which seriously affects the quality of life and survival of patients.With the development of nuclide therapy technology in recent years,223-Ra,as a new type of alpha-targeted therapy,has shown good efficacy in the treatment of desmoplasia-resistant prostate cancer bone metastasis.The purpose of this pa-per is to review the characteristics,mechanism of action,treatment,and the main research results of its treatment of desmoplasia-resistant prostate cancer bone metastasis,and provide a comprehensive review of the clinical application of 223-Ra in the treatment of desmoplasia-resistant prostate cancer bone metastasis for the clinical application of 223-Ra in prostate cancer bone metastasis.
4.Application of microarray chemiluminescent protein chip assay in the diagnosis of systemic lupus erythematosus and comparison with immunoblotting
Yuxuan CHEN ; Wei SHEN ; Shuai DING ; Yang HANG ; Hongxia WEI ; Yue TAO ; Yijia ZHU ; Qisi ZHENG ; Weihua PAN ; Lingyun SUN
Chinese Journal of Rheumatology 2025;29(10):820-829
Objective:To compare the consistency of microarray chemiluminescent protein chip and immunoblotting in the autoantibody spectrum of patients and the diagnostic efficacy of systemic lupus erythematosus(SLE), and to explore the correlation between the detection results of protein microarray and clinical indicators and lymphocyte subsets.Methods:Serum autoantibodies in 649 samples collected between December 2023 and December 2024 in Nanjing Drum Tower Hospital were analyzed using the microarray chemiluminescent protein chip method, with 401 samples simultaneously tested by immunoblotting. Kappa coefficient assessed inter-method consistency. Diagnostic performance was compared via ROC curves. Spearman correlation analysis evaluated relationships between autoantibody levels and SLEDAI-2000 scores, clinical parameters, and lymphocyte subsets.Results:The two methods demonstrated good consistency across 14 antinuclear antibodies, with optimal agreement for anti-SSA/Ro ( Kappa=0.845, P<0.001), anti-SSB ( Kappa=0.825, P<0.001), and anti-CENP B ( Kappa=0.851, P<0.001). The protein chip method significantly improved SLE diagnostic efficacy, particularly for anti-dsDNA (AUC difference=0.291, P<0.001) and anti-Sm antibodies (AUC difference=0.295, P<0.001). Combined detection of anti-SSA/Ro and anti-nRNP/Sm antibodies achieved superior diagnostic performance (AUC=0.927). Anti-dsDNA, anti-histone, and anti-nucleosome antibodies positively correlated with SLEDAI-2000 ( r=0.408, 0410, 0.384, all P<0.001), complement ( P<0.001), and 24-hour urinary protein ( r=0.374, 0.387, 0.301, all P<0.001). Immunological analysis showed that the proportion of NK cells was generally negatively correlated with antinuclear antibodies such as anti-dsDNA ( r=-0.352, P<0.001) and anti-Sm ( r=-0.328, P<0.001) antibodies. Meanwhile, the proportion of CD8 + T cells was positively correlated with anti-nRNP/Sm ( r=0.229, P=0.002) and anti-Sm antibodies ( r=0.211, P=0.005). The proportion of CD4 + T cells was negatively correlated with anti-SSA/Ro ( r=-0.239, P<0.001), while the proportion of B cells was positively correlated with anti-dSDNA antibody ( r=0.300, P<0.001). Conclusion:The protein chip method showed strong consistency with immunoblotting for detecting 14 autoantibodies but demonstrated superior SLE diagnostic efficacy. The combined use of multiple detection methods can enhance the reliability of clinical diagnosis.
5.Organizational Agility in Public Hospitals:Concepts,Core Capabilities,and Development Pathways
Huiyan MAO ; Xuemei WANG ; Jianing WANG ; Yating CHENG ; Yuxuan ZHU ; Tingshan DUAN ; Tao SUN
Chinese Hospital Management 2025;45(3):50-54
In the current era filled with change and challenges,public hospitals,as a crucial part of the national healthcare system,urgently need to enhance their agility to swiftly respond to the ever-changing environment.Current it outlines the origins and connotations of organizational agility,argues for the necessity of organizational agility in public hospitals,and proposes the core competencies required to maintain organizational agility in public hospitals,namely:patient-centeredness,environmental sensitivity,proactive planning,organizational integration capability,flexibility,rapid response,iterative regulation,and continuous learning.Furthermore,it attempts to establish a cultivation pathway for organizational agility in public hospitals,encompassing multiple dimensions such as organizational culture,agile leadership,communication systems,organizational structure,the embedding of new productivity,management systems,partner management,training systems,performance evaluation,and compliance management.
6.Exploration on the Academic Thoughts of Chen Shiduo in the Differentiation and Treatment of Hyperhidrosis
Shuai LIANG ; Luyang LIN ; Yuxuan ZHOU ; Li'an LIU ; Kailong ZHOU ; Qiyu WEN ; Shen'ao TIAN ; Huiyuan SUN
Chinese Journal of Information on Traditional Chinese Medicine 2025;32(6):178-182
This article introduced the experience of physician Chen Shiduo of the Qing Dynasty in treating hyperhidrosis as documented in Bian Zheng Lu.Chen Shiduo proposed several syndromes for hyperhidrosis,including insufficient lung qi,imbalance between the heart and kidney,depletion of kidney yin,excessive stomach fire,and deficiency of heart blood.In terms of treatment,he advocated for methods that focus on reinforcing deficiency and consolidating the exterior while also nourishing yin,clearing heart heat and nourishing kidney water,replenishing true yin to harmonize yin and yang,nourishing stomach yin to reduce fire and stop sweating,and nourishing the heart to enrich blood and drain fire.His medication approach emphasizes harmonizing the internal organs,with a particular expertise in using Mori Folium.His diagnosis and treatment of hyperhidrosis have been remarkably effective and serve as a reference for later generations.
7.Selection and application of pain assessment tools for children
Yuxuan WANG ; Tao SUN ; Hongli ZHENG ; Yu QIAO ; Zhijian FU ; Junnan WANG ; Xiao'en BIAN ; Jing GAO ; Yang CHEN
Chinese Journal of General Practitioners 2025;24(5):613-622
Pain assessment in children is vital in clinical practice. Accurate evaluation of pain intensity in children is the prerequisite for implementing effective analgesic interventions, it is necessary to chose age-specific assessment tools tailored to developmental stages of children. The degrees of patin reported by children themselves are the gold standard for evaluation, and self-assessment should be prioritized when feasible. In addition, behavioral and physiological assessments also show good reliability and validity. This review summarizes current domestic and international researches on pediatric pain assessment tools and their clinical applications, aiming to provide reference for optimizing pain evaluation in pediatric practice.
8.External dose to caregivers and the public during 177Lu-PRRT for neuroblastoma
Yingying SUN ; Di ZUO ; Fei ZHENG ; Han WANG ; Hao YANG ; Yuxuan LIU ; Xiaorong SUN
Chinese Journal of Radiological Medicine and Protection 2025;45(3):216-222
Objective:To estimate the radiation dose to caregivers and the public from 177Lu- peptide receptor radionuclide therapy (PRRT) for pediatric neuroblastoma patients and determine the duration of contact restrictions, in order to provide a reference for relevant radiation protection measures. Methods:A retrospective study was conducted by collecting data from 18 pediatric neuroblastoma patients, aged between 3 and 13 (6.72±2.72), who received 177Lu-1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetic acid-D-Phe1-Tyr3-Thr8-octreotide (DOTATATE) treatment at the Nuclear Medicine Department of Shandong Cancer Hospital from June 2023 to July 2023. Absorbed dose rate in air at 0, 0.1, 0.5, 1 and 2 m from the patients was measured using a radiation-survey at 1, 4, 24, 48 and 96 h after administration. The whole-body region of interest was delined using HERMES software. Subsequently, curve regression fitting was performed using a biexponential function model. By incorporating hypothesized social contact durations, the effective doses received by family members and the public in contact with patients were estimated. Additionally, MIM software was used to outline the whole-body VOI to obtain the total volume of lesions, and the Pearson or Spearman correlation coefficient was employed to analyze the relationship between the absorbed dose rate in air and clinical indicators as well as the total volume of lesions. Results:The 177Lu-DOTATATE administration dose was (4 353.42±1 451.51) MBq. All patients were discharged from hospital 24 h after 177Lu-DOTATATE administration. At the time of discharge, patients had excreted (76.70±3.99)% of the administered activity, and the absorbed dose rate in air at 0.1, 1 and 2 m from the patients were (32.74±6.98), 3.68(3.01, 4.70) and (1.22±0.51) μSv/h, respectively. After being discharged, the radiation doses to caregivers from children aged 2-5 years and 5-13 years were (2.47±1.80) mSv and (0.88±0.47) mSv, respectively. The contact restriction duration was 2 d for nighttime sleeping with family members and 1 d for contact with other children. On the day of discharge, patients should limit their time on public transportation within 4 h and do not need to restrict private transportation. Conclusions:To ensure the effective dose kept within the safety limits stipulated by current regulations, it is necessary to implement contact restrictions for patients’ family members and the public. After implementing preventive measures, 177Lu-DOTATATE treatment is a safe radionuclide therapeutic option.
9.Relationship between aldosterone synthase expression levels in adrenal tissue and prognosis in primary aldosteronism patients
Zhipeng SUN ; Baoan HONG ; Xuezhou ZHANG ; Yuxuan WANG ; Wei WANG ; Yuxuan BO ; Qi MIU ; Mingchuan LI ; Shanshan GONG ; Wei YU ; Dong CHEN ; Ning ZHANG
Chinese Journal of Urology 2025;46(4):241-248
Objective:To investigate the prognostic value of aldosterone synthase (CYP11B2) immunohistochemical expression in adrenal specimens for surgical outcomes of primary aldosteronism (PA).Methods:The clinical data of 99 patients who underwent total unilateral adrenalectomy from June 2022 to January 2023 at Beijing Anzhen Hospital was retrospectively analysed. The clinical data of 99 patients who underwent unilateral total adrenalectomy at Beijing Anzhen Hospital from June 2022 to January 2023 were retrospectively analyzed.There were 59 patients in the PA group, age (53.02±10.56) years, body mass index (BMI) (26.28±4.33) kg/m 2, preoperative aldosterone 29.0(15.9, 61.5)ng/dl, plasma renin 1.3(0.6, 2.8)μIU/ml, aldosterone renin ratio (ARR) 19.3(9.1, 59.2) μg/μIU, preoperative potassium (3.60±0.69) mmol/L, and systolic blood pressure (156.54±21.39) mmHg (1 mmHg=0.133 kPa).There were 40 cases in the nonfunctioning adenoma (NFA) group, age (57.23±9.39) years, BMI (27.07±3.46) kg/m 2, preoperative aldosterone 9.0(7.2, 14.1) ng/dl, plasma renin 18.0(5.2, 47.6)μIU/ml, ARR 0.6(0.2, 1.4) μg/μIU, preoperative potassium (4.17±0.41) mmol/L, and systolic blood pressure (157.97±26.87) mmHg. The differences between the two groups were statistically significant for potassium ( P<0.01), aldosterone ( P=0.012), renin ( P<0.01), and ARR ( P<0.01).Surgical outcomes were assessed using the Consensus on the Outcome of Surgery for Primary Aldosteronism (PASO) (complete/partial/no success for clinical and biochemical outcomes). CYP11B2 expression was evaluated by immunohistochemistry using the 2022 World Health Organization's histopathology of primary aldosteronism (HISTALDO) criteria. The correlation between the expression of CYP11B2 and surgical outcomes was assessed. Results:The mean follow-up of 99 patients was (11.73±4.92) months. Of these, 36 out of 59 PA patients had positive CYP11B2 expression in their adrenal specimens, while 23 were negative; all 40 NFA patients were negative for CYP11B2. Among the 36 CYP11B2-positive PA patients, there were 19 cases of aldosterone-producing adenomas, 3 aldosterone-producing nodules, 4 aldosterone-producing micronodules, 8 multiple aldosterone-producing micronodules, and 2 aldosterone-producing diffuse hyperplasia. 36 cases of CYP11B2-positive PA patients had complete clinical success in 15 cases, partial success in 20 cases, and no success in 1 case, and complete biochemical success in 24 cases, partial success in 11 cases, and no success in 1 case; 23 CYP11B2-negative PA patients had complete clinical success in 4 cases, partial success in 15 cases, and no success in 4 cases, and complete biochemical success in 6 cases, partial success in 15 cases, and no success in 2 cases. Adrenal specimens from CYP11B2-positive PA patients had significantly better clinical ( P=0.038) and biochemical ( P=0.008) success rates than CYP11B2-negative PA patients. Patients with aldosterone-producing adenomas had complete clinical success in 8 cases, partial success in 11 cases, and no success in 0 cases, and biochemical success was completely achieved in 16 cases, partially achieved in 2 cases, and not successful in 1 case. They also had significantly higher clinical ( P=0.028) and biochemical ( P<0.01) success rates compared to CYP11B2-negative PA patients. Conclusions:Patients with PA who had immunohistochemical staining for CYP11B2 positivity and high expression in adrenal specimens had a better postoperative clinical and biochemical prognosis. Patients with aldosterone-producing adenomas had the greatest postoperative outcome of all pathological subtypes of PA.
10.Identification of natural product-based drug combination (NPDC) using artificial intelligence.
Tianle NIU ; Yimiao ZHU ; Minjie MOU ; Tingting FU ; Hao YANG ; Huaicheng SUN ; Yuxuan LIU ; Feng ZHU ; Yang ZHANG ; Yanxing LIU
Chinese Journal of Natural Medicines (English Ed.) 2025;23(11):1377-1390
Natural product-based drug combinations (NPDCs) present distinctive advantages in treating complex diseases. While high-throughput screening (HTS) and conventional computational methods have partially accelerated synergistic drug combination discovery, their applications remain constrained by experimental data fragmentation, high costs, and extensive combinatorial space. Recent developments in artificial intelligence (AI), encompassing traditional machine learning and deep learning algorithms, have been extensively applied in NPDC identification. Through the integration of multi-source heterogeneous data and autonomous feature extraction, prediction accuracy has markedly improved, offering a robust technical approach for novel NPDC discovery. This review comprehensively examines recent advances in AI-driven NPDC prediction, presents relevant data resources and algorithmic frameworks, and evaluates current limitations and future prospects. AI methodologies are anticipated to substantially expedite NPDC discovery and inform experimental validation.
Artificial Intelligence
;
Biological Products/chemistry*
;
Humans
;
Drug Combinations
;
Drug Discovery/methods*
;
Machine Learning
;
Algorithms

Result Analysis
Print
Save
E-mail