Adverse drug reactions (ADRs) significantly impact clinical medication safety. The timely identification and prediction of ADRs rely on the efficient analysis of real-world data, such as electronic health records, social media, and spontaneous reporting databases. In recent years, the rapid advancement of artificial intelligence, particularly large language models, in natural language processing, causal reasoning, and complex data mining has provided new technological means for real-time ADRs monitoring and individualized prediction. This paper summarizes the latest research achievements in AI-driven ADRs monitoring. Focusing on diverse data sources, including structured databases and electronic health records, it elaborates on the advantages andchallenges of AI in ADRs event extraction, relationship identification, causal analysis, and risk prediction. The aim is to provide a theoretical reference for constructing more intelligent and efficient ADRs monitoring systems.

Objective To analyze dynamic contrast-enhanced MRI (DCE-MRI) radiomic features using machine learning algorithms, and to develop and validate a predictive model for HER-2 status in breast cancer. Methods The DCE-MRI images of 272 treatment-naive female patients with breast cancer between 2020 and 2022 were included in this study. Regions of interest (ROIs) were manually segmented using 3d-Slicer software, and radiomic features were extracted. All patients were randomly divided into training sets or validation sets at a ratio of 4∶1. The least absolute shrinkage and selection operator (LASSO) algorithm was used for feature screening on the training set, followed by the development of predictive models using six machine learning algorithms. Internal cross-validation was performed to compare the performance differences between the models. The best-performing model was selected, trained on the training set, and evaluated on the validation set. Evaluation metrics included area under the curve (AUC), sensitivity, specificity, precision, and recall rate. Results The clinical data of patients in the training set and validation set showed no significant differences. Five features were identified by the LASSO algorithm. With these features, six machine learning models were developed on the training set, and their predictive performance was internally cross-validated using the bagging method. XGBoost model had the highest mean AUC (0.696), followed by RF model (0.690); XGBoost model had the highest mean precision (0.756), followed by LR and RF models. Therefore, XGBoost was the optimal model. An HER-2 predictive model was built using the XGBoost algorithm on the training set and applied to the validation set. The AUC, precision, sensitivity, and specificity of the predictive model on the validation set were calculated, and ROC curves, precision-recall curves, calibration curves, and decision-making curves were plotted. Conclusion This study constructed and evaluated different DCE-MRI radiomics-based machine learning models for predicting HER-2 status in breast cancer. Among them, XGBoost algorithm performed the best and has the potential to become a new non-invasive method for preoperative prediction of HER-2 status, providing reliable evidence for personalized clinical diagnosis and treatment.

Objective:To explore the relationship between general demographic characteristics,inflammatory indicators,nutritional indicators,pathological data and lymph node metastasis in endometrial cancer(EC)pa-tients,and to construct and validate a model for preoperative prediction of lymph node status in endometrial canc-er patients.Methods:The preoperative clinical data of 473 patients with EC who underwent surgical treatment in the Zhu Jiang Hospital of Southern Medical University from January 2010 to April 2024 were retrospectively ana-lyzed.The independent risk factors of lymph node metastasis of endometrial cancer were screened by univariate and multivariate Logistic regression analyses,and the nomogram prediction model was constructed by R soft-ware.The performance of the model was evaluated by the receiver operating characteristic(ROC)curve,calibra-tion curve and clinical decision curve.Results:Menopausal status,high grade biopsy pathology,CA125 ≥24.47U/ml,systemic immune inflammatory index(SII)≥710.91,and prognostic nutritional index(PNI)＜52.90 were in-dependent risk factors for lymph node metastasis in endometrial cancer(OR＞1,P＜0.05).The nomogram model constructed based on these five factors had an AUC of 0.853 in the training set and 0.871 in the test set.The cali-bration curve fitted well,and the clinical decision curve shows a positive benefit.Conclusions:The endometrial cancer lymph node metastasis prediction model constructed based on menopausal status,biopsy pathology,CA125,SII,and PNI has good accuracy and fit,with certain clinical application value.

Alzheimer's disease(AD)is the leading cause of dementia,and no effective treatment has been devel-oped for it thus far.Recently,the use of natural compounds in the treatment of neurodegenerative diseases has garnered significant attention owing to their minimal adverse reactions.Accordingly,the potential therapeutic effect of pterostilbene(PTS)on AD has been demonstrated in multiple in vivo and in vitro experiments.In this study,we systematically reviewed and summarized the results of these studies investigating the use of PTS for treating AD.Analysis of the literature revealed that PTS may play a role in AD treatment through various mechanisms,including anti-oxidative damage,anti-neuroinflammation,anti-apoptosis,cholinesterase activity inhibition,attenuation of β-amyloid deposi-tion,and tau protein hyperphosphorylation.Moreover,PTS interferes with the progression of AD by regulating the activities of peroxisome proliferator-activated receptor alpha(PPAR-α),monoamine oxi-dase B(MAO-B),silent information regulator sirtuin 1(SIRT1),and phosphodiesterase 4A(PDE4A).Furthermore,to further elucidate the potential therapeutic mechanisms of PTS in AD,we employed network pharmacology and molecular docking technology to perform molecular docking of related proteins,and the obtained binding energies ranged from-2.83 to-5.14 kj/mol,indicating that these proteins exhibit good binding ability with PTS.Network pharmacology analysis revealed multiple po-tential mechanisms of action for PTS in AD.In summary,by systematically collating and summarizing the relevant studies on the role of PTS in treatment of AD,it is anticipated that this will serve as a reference for the precise targeted prevention and treatment of AD,either using PTS or other developed drug interventions.

Alzheimer's disease (AD) is the leading cause of dementia, and no effective treatment has been developed for it thus far. Recently, the use of natural compounds in the treatment of neurodegenerative diseases has garnered significant attention owing to their minimal adverse reactions. Accordingly, the potential therapeutic effect of pterostilbene (PTS) on AD has been demonstrated in multiple in vivo and in vitro experiments. In this study, we systematically reviewed and summarized the results of these studies investigating the use of PTS for treating AD. Analysis of the literature revealed that PTS may play a role in AD treatment through various mechanisms, including anti-oxidative damage, anti-neuroinflammation, anti-apoptosis, cholinesterase activity inhibition, attenuation of β-amyloid deposition, and tau protein hyperphosphorylation. Moreover, PTS interferes with the progression of AD by regulating the activities of peroxisome proliferator-activated receptor alpha (PPAR-α), monoamine oxidase B (MAO-B), silent information regulator sirtuin 1 (SIRT1), and phosphodiesterase 4A (PDE4A). Furthermore, to further elucidate the potential therapeutic mechanisms of PTS in AD, we employed network pharmacology and molecular docking technology to perform molecular docking of related proteins, and the obtained binding energies ranged from -2.83 to -5.14 kJ/mol, indicating that these proteins exhibit good binding ability with PTS. Network pharmacology analysis revealed multiple potential mechanisms of action for PTS in AD. In summary, by systematically collating and summarizing the relevant studies on the role of PTS in treatment of AD, it is anticipated that this will serve as a reference for the precise targeted prevention and treatment of AD, either using PTS or other developed drug interventions.

The classical latent structure method does not consider the influence of primary and secondary symptoms,syndromes and symptoms in the analysis and modeling of syndromes.In this paper,based on the data of damp-heat in intestine and stomach syndrome involving 1087 diabetic patients,the classical latent structure analysis was used to obtain the quantitative syndrome diagnostic rules.Then,using Constrained Latent Tree Analysis(CLTA),the quantitative syndrome diagnostic rules containing primary and secondary symptoms were obtained as follows,primary symptoms include halitosis(2.3),yellow tongue coating(2),abdominal distension(2.3),greasy tongue coating(2.1),loose stool or loose stool(1.5),red tongue(1.3),smooth pulse(1.4).Secondary symptoms include epigastric distension(1.1).Compared with the traditional latent structure analysis method,the rules established by CLTA are more compatible with the concept of differentiating primary and secondary symptoms and the common practice of TCM.The quantitative syndrome diagnostic rules of damp-heat in intestine and stomach syndrome constructed by the CLTA method have considerable objectivity in the modeling process.The diagnostic rules established were also compatible with the qualitative concept of TCM theory in stratifying primary and secondary symptoms.Finally,the diagnostic rules are obtained by logistic regression analysis,and the accuracy of the three rules is compared.The results show that the rule recognition accuracy obtained by CLTA is the highest.Therefore,the syndrome diagnostic rules of damp-heat in intestine and stomach obtained from the analysis of CLTA are in line with the constraint semantics of primary and secondary diseases and the theory of traditional Chinese medicine.

Objective To investigate the influencing factors of pulmonary and extrapulmonary acute re-spiratory distress syndrome(ARDS)caused by sepsis.Methods A total of 126 patients with ARDS admitted to the Department of Critical Care Medicine,General Hospital of Ningxia Medical University,from January 2022 to June 2024 were selected.Patients were divided into pulmonary ARDS and extrapulmonary ARDS groups based on the etiology of ARDS.General data,inflammatory indicators,and prognostic outcomes were compared between the two groups.COX regression analysis was used to identify prognostic factors.Results A-mong the 126 patients,72 were diagnosed with pulmonary ARDS and 54 with extrapulmonary ARDS.The pulmonary ARDS group had significantly lower SOFA scores,fewer organ dysfunctions,a lower incidence of arrhythmia,shorter mechanical ventilation duration,higher Murray scores,and higher Charlson Comorbidity Index(CCI)compared to the extrapulmonary ARDS group(P<0.05).Inflammatory markers,including pro-calcitonin(PCT),C-reactive protein(CRP),interleukin(IL)-4,IL-6,IL-10,and tumor necrosis factor-α(TNF-α),were significantly lower in the pulmonary ARDS group,while interferon-γ(INF-γ)levels were higher(P<0.05).For pulmonary ARDS,CCI and TNF-α were identified as independent risk factors for prog-nosis(P<0.05),with the combination of CCI and TNF-α yielding the highest predictive accuracy(AUC=0.81,95%CI:0.71-0.91).For extrapulmonary ARDS,CCI and CRP were independent risk factors(P<0.05),and their combination achieved the highest predictive performance(AUC=0.91,95%CI:0.84-0.98).Conclusion Inflammatory profiles between pulmonary and extrapulmonary ARDS caused by sepsis are different.CCI and TNF-α are independent risk factors for mortality in pulmonary ARDS,while CCI and CRP are independent risk factors in extrapulmonary ARDS.

Objective:To explore the efficacy of non-invasive prenatal testing (NIPT) of fetal free DNA in maternal peripheral blood in fetuses with increased nuchal translucency (NT).Methods:A retrospective analysis was conducted on 1 184 singleton pregnant women that underwent chromosomal microarray analysis (CMA) at Nanjing Drum Tower Hospital, Nanjing University Medical School from June 2014 to December 2022 due to fetal increased NT (≥3.0 mm). These subjects were categorized based on whether the increased NT was accompanied by other high-risk factors into isolated increased NT without advanced maternal age (further subdivided into 3.0 mm≤NT<3.5 mm, 3.5 mm≤NT<4.0 mm, and NT≥4.0 mm subgroups), isolated increased NT with advanced maternal age, increased NT with nasal bone abnormalities, increased NT with other soft markers, and increased NT with structural abnormalities groups. Assuming the sensitivity and specificity of NIPT and expanded NIPT at this center were both 100%, genomic abnormalities outside the detection range of NIPT or expanded NIPT were termed as residual risk of NIPT or expanded NIPT. Chi-square test and Bonferroni correction were used to compare the residual risks of NIPT and expanded NIPT among the three subgroups of isolated increased NT without advanced maternal age group. Results:(1) In the group of isolated increased NT without advanced maternal age: For the 3.0 mm≤NT<3.5 mm subgroup (329 cases), 19 abnormalities were detected by CMA [12 cases of chromosome aneuploidy, seven cases of pathogenic copy number variation (pCNV)], with residual risks of NIPT and expanded NIPT both at 2.1% (7/329). For the 3.5 mm≤NT<4.0 mm subgroup (173 cases), 29 abnormalities were detected by CMA (17 cases of chromosome aneuploidy, nine cases of pCNV, three cases of chromosome unbalanced translocation), with residual risks of NIPT at 8.1% (14/173) and expanded NIPT at 7.5% (13/173). For the NT≥4.0 mm subgroup (270 cases), CMA detected abnormalities in 70 cases (50 cases of chromosome aneuploidy, 16 cases of pCNV, three cases of unbalanced translocations, and one case of sex chromosome abnormality combined with pCNV). The residual risk of NIPT was 12.2% (33/270), and the residual risk of expanded NIPT was 7.0% (19/270). The residual risks of NIPT and expanded NIPT in the 3.0 mm≤NT<3.5 mm subgroup were lower than those in the 3.5 mm≤NT<4.0 mm and NT≥4.0 mm subgroups (Bonferroni correction, all P<0.017). (2) In the group of 92 cases with isolated increased NT and advanced maternal age, CMA detected abnormalities in 36 cases (29 cases of chromosome aneuploidy, five cases of pCNV, one case of trisomy 21 combined with sex chromosome abnormality, and one case of trisomy 18 combined with sex chromosome abnormality). The residual risk of NIPT was 7.6% (7/92), and that of expanded NIPT was 5.4% (5/92). (3) In the group of 49 cases with increased NT combined with nasal bone abnormalities, CMA detected abnormalities in 24 cases (23 cases of chromosome aneuploidy and one case of pCNV). The residual risks of NIPT and expanded NIPT were both 2.0% (1/49). (4) In the group of 26 cases with increased NT combined with other soft markers, CMA detected abnormalities in nine cases (six cases of chromosome aneuploidy, one case of pCNV, and two cases of chromosome unbalanced translocations). The residual risks of NIPT and expanded NIPT were both 11.5% (3/26). (5) In the group of 245 cases with increased NT combined with structural abnormalities, CMA detected abnormalities in 121 cases (107 cases of chromosome aneuploidy, seven cases of pCNV, four cases of chromosome unbalanced translocations, one case of trisomy 21 combined with trisomy 20, and two cases of trisomy 18 combined with sex chromosome abnormalities). The residual risk of NIPT was 16.7% (41/245), and that of expanded NIPT was 4.1% (10/245). Conclusions:For isolated NT≥3.5 mm or NT≥3.0 mm combined with other high-risk factors, chorionic villus sampling in early pregnancy can be recommended, advancing the timing of prenatal diagnosis from the second trimester to the first trimester. For fetuses with isolated 3.0 mm≤NT<3.5 mm, the 2.1% residual risk of chromosomal abnormalities should be fully informed during counseling, even if the risk of NIPT is low.

As a major global public health concern,cancer has witnessed a continues rise in both incidence and mortality rates.It pose not only a severe threat to human health but also a heavy burden on socioeconomic systems.Despite remarkable advancements in oncology research,critical challenges such as tumor heterogeneity,drug resistance,and limitations in early screening and diagnostic technologies remain to be addressed.Against this backdrop,artificial intelligence(AI),with its unique advantages in big data analysis,pattern recognition,and predictive modeling,has opened new avenues for cancer research.By integrating multi-modal data,including omics,imaging,and clinical information,AI not only accelerates investigations into fundamental tumor mechanisms but also demonstrates immense potential in areas such as early screening,biomarker discovery,and personalized treatment.These advancements have fostered a deeper integration of precision medicine and oncology.This review provides a comprehensive overview of the most recent progresses in the application of AI in cancer diagnosis and treatment research,with a focus on its practical value across diverse data types and clinical scenarios,as well as future directions for its development.

Objective:To analyze the effect and safety of tirofiban bridging combined with aspirin for anti-platelet therapy in patients with acute cerebral infarction (ACI) beyond the thrombolytic time window.Methods:Sixty patients with ACI treated in Beijing Royal Hospital from January 2021 to September 2022 were retrospectively analyzed. Among them, 30 cases were treated with tirofiban bridging aspirin as the observation group, while 30 cases were treated with aspirin alone as the control group. The total effective rate of clinical treatment, the degree of neurological deficit, the degree of disease outcome and the incidence of adverse reactions were compared between the two groups. The degree of neurological deficit was evaluated using the National Institutes of Health stroke scale (NIHSS), and the degree of disease outcome was evaluated using the modified Rankin scale (mRS).Results:The overall effective rate in the observation group was significantly higher than that in the control group: 93.3% (28/30) vs. 66.7% (20/30), with a statistically significant difference ( χ2 = 10.97, P<0.01). Before treatment, there was no statistically significant difference in NIHSS score between the two groups ( P>0.05). However, after treatment, the NIHSS score in the observation group was significantly lower than that in the control group: (2.83 ± 1.87) scores vs. (4.93 ± 3.05) scores, indicating a significant difference ( t = -3.21, P = 0.002). No significant difference was observed in mRS score between the two groups before treatment( P>0.05). After treatment, the observation group showed significantly lower mRS score compared to the control group: 2.00 (1.00, 2.00) scores vs. 3.00 (2.00, 4.00) scores ( P = 0.006). There was no statistical difference in the incidence of adverse reactions between the observation group and control groups ( P>0.05). Conclusions:Tirofiban bridging aspirin has stronger inhibitory effect on platelet activity than simple application of aspirin for ACI patients with ultra-thrombolytic window, which can rapidly improve the degree of neurological impairment and daily living ability of patients, and the incidence of adverse reactions has not increased significantly.