1.Explainable Machine Learning Model for Predicting Prognosis in Patients with Malignant Tumors Complicated by Acute Respiratory Failure: Based on the eICU Collaborative Research Database in the United States
Zihan NAN ; Linan HAN ; Suwei LI ; Ziyi ZHU ; Qinqin ZHU ; Yan DUAN ; Xiaoting WANG ; Lixia LIU
Medical Journal of Peking Union Medical College Hospital 2026;17(1):98-108
To develop and validate a model for predicting intensive care unit (ICU) mortality risk in patients with malignant tumors complicated by acute respiratory failure (ARF) based on an explainable machine learning framework. Clinical data of patients with malignant tumors and ARF were extracted from the eICU Collaborative Research Database in the United States, including demographic characteristics, comorbidities, vital signs, laboratory test indicators, and major interventions within the first 24 hours after ICU admission.The study outcome was ICU death.Enrolled patients were randomly divided into a training set and a validation set at a ratio of 7:3.Predictor variables were selected using least absolute shrinkage and selection operator (LASSO) regression.Five machine learning algorithms-extreme gradient boosting (XGBoost), support vector machine (SVM), Logistic regression, multilayer perceptron (MLP), and C5.0 Decision Tree-were employed to construct predictive models.Model performance was evaluated based on the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and other metrics.The optimal model was further interpreted using the Shapley additive explanations (SHAP) algorithm. A total of 3196 patients with malignant tumors complicated by ARF were included.The training set comprised 2, 261 patients and the validation set 935 patients; 683 patients died during ICU stay, while 2513 survived.LASSO regression ultimately selected 12 variables closely associated with patient ICU outcomes, including sepsis comorbidity, use of vasoactive drugs, and within the first 24 hours after ICU admission: minimum mean arterial pressure, maximum heart rate, maximum respiratory rate, minimum oxygen saturation, minimum serum bicarbonate, minimum blood urea nitrogen, maximum white blood cell count, maximum mean corpuscular volume, maximum serum potassium, and maximum blood glucose.After model evaluation, the XGBoost model demonstrated the best performance.The AUCs for predicting ICU mortality risk in the training and validation sets were 0.940 and 0.763, respectively; accuracy was 88.3% and 81.2%;sensitivity was 98.5% and 95.9%.Its predictive performance also remained optimal in sensitivity analyses.SHAP analysis indicated that the top five variables contributing to the model's predictions were minimum oxygen saturation, minimum serum bicarbonate, minimum mean arterial pressure, use of vasoactive drugs, and maximum white blood cell count. This study successfully developed a mortality risk prediction model for ICU patients with malignant tumors complicated by ARF based on a large-scale dataset and performed explainability analysis.The model aids clinicians in early identification of high-risk patients and implementing individualized interventions.
2.Exploring Chemical Constituent Distribution in Blood/Brain(Hippocampus) and Emotional Regulatory Effect of Raw and Vinegar-processed Products of Citri Reticulatae Pericarpium Viride
Yi BAO ; Yonggui SONG ; Qianmin LI ; Zhifu AI ; Genhua ZHU ; Ming YANG ; Huanhua XU ; Qin ZHENG ; Yiting HUANG ; Zihan GAO ; Dan SU
Chinese Journal of Experimental Traditional Medical Formulae 2026;32(2):189-197
ObjectiveTo investigate the migration and distribution characteristics of chemical constituents in blood and hippocampal tissues before and after vinegar processing of Citri Reticulatae Pericarpium Viride(CRPV), and to explore the potential material basis and mechanisms underlying their regulatory effects on emotional disorders by comparing the effects of raw and vinegar-processed products of CRPV. MethodsUltra-performance liquid chromatography-quadrupole-time-of-flight mass spectrometry(UPLC-Q-TOF-MS/MS) was employed to characterize and identify the chemical constituents of raw and vinegar-processed products of CRPV extracts, as well as their migrating components in blood and hippocampal tissues after oral administration. Reference standards, databases, and relevant literature were utilized for compound annotation, with data processing performed using PeakView 1.2 software. Seventy male C57BL/6 mice were randomly divided into seven groups, including the blank group, model group, diazepam group(2.5 mg·kg-1), raw CRPV low/high dose groups(0.6, 1.2 g·kg-1), and vinegar-processed CRPV low/high dose groups(0.6, 1.2 g·kg-1), with 10 mice per group. Except for the blank group, all other groups underwent chronic restraint stress(2 h·d-1) for 20 d. Each drug-treated group received oral administration at the predetermined dose starting 10 d after modeling, with a total treatment duration of 10 d. Following model-based drug administration, mice underwent open-field, forced swimming, and elevated plus maze tests. After anesthesia with isoflurane, whole brains were collected from each group of mice, and hippocampi were dissected. Reactive oxygen species(ROS) level in hippocampal tissues was quantified by enzyme-linked immunosorbent assay(ELISA). Hematoxylin-eosin(HE) staining was used to observe hippocampal tissue morphology. Immunofluorescence was performed to detect neuronal nuclei(NeuN) and peroxisome proliferator-activated receptor alpha(PPARα) expressions in hippocampal tissue. Then, pharmacodynamic evaluations were conducted to assess the effects of raw and vinegar-processed CRPV on mood disorders, exploring the potential mechanisms. ResultsVinegar processing caused significant changes in the chemical composition of CRPV, with 18 components showing increased relative content and 35 components showing decreased relative content. The primary changes occurred in flavonoid compounds, including 20 flavonoids, 20 flavonoid glycosides, 3 triterpenes, 3 phenolic acids, 1 alkaloid, and 6 other compounds. Twenty-one components were detected in blood(15 methoxyflavones, 4 flavonoid glycosides, and 2 phenolic acids), with 17 shared between raw and vinegar-processed CRPV. Seven components reached hippocampal tissues(all common to both forms). In regulating emotional disorders, Vinegar-processed CRPV exhibited superior antidepressant-like effects compared to raw products. HE staining revealed that both treatments improved hippocampal neuronal morphology, particularly in the damaged CA1 and CA3 regions. Immunofluorescence and ELISA analyses demonstrated that both raw and vinegar-processed CRPV significantly modulated NeuN and PPARα expressions in hippocampal tissue while alleviating oxidative stress induced by excessive ROS(P<0.05). ConclusionThe chemical composition of CRPV undergoes changes after vinegar processing, but the migrating components in blood and hippocampus are primarily methoxyflavonoids. These components may serve as the potential material basis for activating the PPARα pathway, thereby negatively regulating ROS generation in the hippocampus, reducing oxidative stress, and promoting the development of NeuN-positive neurons. These findings provide experimental evidence for enhancing quality standards, pharmacodynamic material research, and active drug development of raw and vinegar-processed CRPV.
3.The current status of international health communication research and its implications for China
Lingyan YANG ; Zihan YU ; Yueqiao ZHAO ; Zhenping LI ; Jianyi YAO ; Hao LI ; Yuhui ZHOU
Journal of Public Health and Preventive Medicine 2026;37(1):18-21
Objective To systematically review international research on health communication, and to provide valuable insights and reference for China's health communication research and practice. Methods This study included 693 articles published from January 2023 to April 2024 in two authoritative academic journals in the field of health communication, “Health Communication” and the “Journal of Health Communication”. A systematic review was conducted on the themes, theoretical foundations, research methods, and populations of international health communication research. Results The findings in this study revealed that international health communication research topics were diverse, with hotspots including social media, health information behavior, health misinformation, stigmatization, trust, and risk perception. The results showed that 34% of the articles were based on theoretical foundations, and 93.3% employed research methods, focusing on adolescents, parents, women, and other key populations. Conclusion Domestic health communication research can expand its perspective from “information transmission” to “social interaction”, innovate theories and methods from “single paradigm" to “multi-integration” and shift focus from a “mass perspective” to “targeted care” for the health of all populations. Domestic health communication practice can delve into the localization of social media health communication practices, the comprehensive management of health misinformation, and the critical application of new technologies.
4.Peyton's Four-Step Teaching Method for Intestinal Ultrasound Training: Efficacy and Practical Implications
Zihan NIU ; Xiaoyan ZHANG ; Zhaojue WANG ; Qingli ZHU ; Mengsu XIAO ; Li MA ; Yudi HE ; Wenbo LI
Medical Journal of Peking Union Medical College Hospital 2026;17(2):591-596
To evaluate the application value of the Peyton four-step teaching method in the standardized training of intestinal ultrasound and compare it with traditional teaching methods, so as to provide an optimized approach for clinical ultrasound training. Participants from the Department of Ultrasound at Peking Union Medical College Hospital between September 2024 and March 2025 were randomly assigned to either the traditional group or Peyton group. The traditional group followed the conventional "lecture- demonstration-practice" model, while the Peyton group implemented the standardized "demonstration-deconstruction-comprehension-execution" four-step approach. All training focused on standard intestinal ultrasound scanning techniques. After the training, the operational skills were independently evaluated by the instructors. To verify the reproducibility of the teaching method, the participants in traditional teaching group received additional Peyton method training after the initial assessment and underwent a second evaluation. A total of 18 participants were included in this study, with 9 in the traditional teaching group and 9 in the Peyton teaching group. Participants in the Peyton group demonstrated significantly higher scores than those in the traditional group at every anatomical site assessed (all The Peyton four-step method is significantly more effective than traditional teaching in improving residents' intestinal ultrasound skills, demonstrating its suitability as the preferred approach for standardized training programs.
5.Comparative Analysis of Clinical Efficacy of Traditional Chinese Medicine Manipulative Reduction Combined with Small Splint Fixation Versus Surgical Treatment for Type A Distal Radius Fracture
Yang SHAO ; Zihan WANG ; Jianwei WANG ; Guoda DAI ; Hengyan CUI ; Zhen HUA ; Tingchen ZHU ; Shaoshuo LI ; Jun MAO ; Fenghua CHEN ; Shuai TAO ; Mao WU
Journal of Traditional Chinese Medicine 2026;67(10):1078-1085
ObjectiveTo compare the clinical efficacy of traditional Chinese medicine (TCM) manipulative reduction combined with small splint fixation versus surgical treatment for type A distal radius fracture (DRF) and to explore the factors influencing the choice of treatment. MethodsA multi-center retrospective study was conducted, collecting data from 1237 type A DRF patients treated in 11 hospitals in Jiangsu province from September, 2023 to April, 2025. Among them, 851 patients in the TCM group received manipulative reduction combined with small splint fixation, and 386 patients in the surgical group underwent open reduction and internal fixation. Visual analog scale (VAS) scores for pain and radiographic indicators including palmar tilt, ulnar deviation, and radial height were compared before treatment, 5-7 days after treatment, and 4-6 weeks after treatment. The wrist joint function scores including Dienst and Gartland-Werley scores at 12 weeks after treatment were recorded. Subgroup analysis was conducted for the excellent rate of Dienst and Gartland-Werley scores, stratified by age (<50, 50-59, 60-69, ≥70 years old) and AO subtypes (A1, A2, A3). A multivariate logistic regression model was used to identify independent factors influencing treatment choice. ResultsOn 5-7 days after treatment, the surgical group had lower VAS scores than the TCM group, while 4-6 weeks after treatment, the TCM group showed lower VAS scores than the surgical group (P<0.01). In terms of radiographic indicators, except for the palmar tilt before treatment being higher in the surgical group than in the TCM group (P<0.01), there were no significant differences in palmar tilt, ulnar deviation, and radial height at other timepoints (P>0.05). Twelve weeks after treatment, the surgical group had a higher average Gartland-Werley score and the excellent rate than the TCM group (P<0.01). Subgroup analysis showed that in patients with A2 type DRF aged 50-59 and 60-69 years old, the excellent rates of Dienst and Gartland-Werley scores in the TCM group were higher than those in the surgical group (P<0.05). Multivariate logistic regression analysis revealed that age, palmar tilt, ulnar deviation, and the degree of swelling on the affected side were independent factors influencing the choice of treatment (P<0.05). ConclusionBoth TCM manipulative reduction combined with small splint fixation and surgical treatment for type A DRF can achieve good therapeutic effects. TCM manipulative reduction combined with small splint fixation has certain advantages in medium- and long-term pain relief, especially in elderly patients, where wrist joint function recovery is more stable. Age, palmar tilt, ulnar deviation, and swelling degree are the main factors influencing the treatment choice.
6.Dissecting antibody-mediated natural killer cell effects reveals a cytotoxic CX3CR1+KLRC2–CD16hi subset linked to hepatitis B virus outcomes
Libo TANG ; Yuhao WANG ; Zihan JIN ; Yurong GU ; Zhaofeng ZENG ; Linnan SONG ; Xuan YI ; Lingtao ZHANG ; Yujing ZHANG ; Weiying HE ; Liping WANG ; Weixin HE ; Jianru SUN ; Xiaoqin LAN ; Xiangyong LI ; Shihong ZHONG ; Yongyin LI
Clinical and Molecular Hepatology 2026;32(2):683-705
Background/Aims:
Natural killer (NK) cell function is generally considered dampened in chronic hepatitis B virus (HBV) infection; however, the NK cell pool exhibits phenotypic and functional heterogeneity, and the antibody--mediated effect of NK cells remains less characterized. This study evaluated the dynamic changes in antibody-mediated NK cell responses and the involvement of distinct NK subsets across disease stages and during antiviral treatment.
Methods:
A T-cell receptor-like antibody specific for the HBV core 18–27 peptide (cTCRL-Ab) was used to determine the antibody-mediated effect of NK cells, and an array of NK cell surface markers were analyzed in cross-sectional and longitudinal cohorts of patients with chronic HBV infection. Single-cell RNA sequencing (scRNA-seq) was performed to identify the heterogeneity of NK subsets.
Results:
The cTCRL-Ab enabled the detection of NK cell cytolytic activity and IFNγ production. Notably, cTCRL-Ab-mediated NK cell responses were compromised in chronically HBV-infected patients, particularly in those receiving pegylated interferon-α (Peg-IFNα), which was associated with the downregulation of CD16 expression. Correspondingly, Peg-IFNα inhibited cTCRL-Ab-mediated NK cell function by reducing CD16 expression in vitro. scRNA-seq revealed that CD16 downregulation occurred mainly within a dysfunctional CD16hi NK subset exhibiting exhaustion properties. In contrast, an activated CD16hiNK subpopulation (CX3CR1⁺KLRC2–CD16hi) with high cytotoxicity was enriched in patients who experienced favorable treatment responses. Furthermore, the intrahepatic CX3CR1+KLRC2–CD16hi subset tended to exhibit functional restoration in HBsAg-loss individuals.
Conclusions
Our data contribute to the understanding of antibody-mediated responses of NK cells in chronic HBV infection, and highlight a previously unappreciated functional CX3CR1+KLRC2–CD16hiNK subset as a potential therapeutic target.
7.The therapeutic effects of MCC950 on benzalkonium chloride-induced dry eye in rats
Zihan LIANG ; Chenlu ZHU ; Kexuan REN ; Xiaofeng LI
Recent Advances in Ophthalmology 2025;45(10):787-792
Objective To investigate the therapeutic effect and mechanism of MCC950 on benzalkonium chloride-in-duced dry eye(DED)in rats.Methods Among 60 male SD rats,50 rats were treated with 20 g·L-1 benzalkonium chlo-ride solution 3 times a day in both eyes to induce DED model,and a significant decrease in tear secretion was regarded as a successful modeling,and the remaining 10 rats were regarded as the control group.DED rats were randomly divided into 5 groups:DED,DED+10MCC950,DED+50MCC950,DED+100MCC950,and DED+500MCC950 groups,and were given 0.2 mL of 0,10,50,100,and 500 μmol·L-1 MCC950 in both eyes for two weeks,and then examined the amount of tear secretion,corneal histopathology(HE staining),reactive oxygen species(ROS),and inflammation factors(NLRP3,Caspase-1,IL-1β,IL-6).Results The tear secretion of rats in the DED group was significantly lower than that of the control group(P<0.05).Except for the DED+10MCC950 group,the tear secretion of rats in different concentrations of MCC950 treatment groups was increased and positively correlated with the MCC950 dose(all P<0.05).HE staining of the cornea showed that compared with the control group,the cornea of the DED group was significantly thinner,the arrange-ment of surface cells was disorganized,the number of cells was significantly reduced,and a large number of"vacuole"-like structures appeared;the corneal stroma was disorganized and sparsely structured.The degree of histopathological changes in the cornea of rats treated with different concentrations of MCC950 decreased with the increase of the concentration of MCC950.Compared with the control group,the relative expression level of ROS in the DED group of rats was significantly increased(P<0.05).In each MCC950 treatment group,as the concentration of MCC950 increased,the relative expression level of corneal ROS staining in rats gradually decreased,showing a concentration-dependent effect(all P<0.05).Com-pared with the control group,the levels of NLRP3,IL-1 β,Caspase-1 and IL-6 were significantly increased in the DED group(all P<0.05).The levels of various inflammatory factors were dose-dependently reduced in the MCC950 treatment groups(all P<0.05).Conclusion MCC950 can reduce oxidative stress and inflammation and improve the symptoms of benza-lkonium chloride-induced DED in rats by inhibiting the activation of NLRP3 inflammatory vesicles.
8.Identification of Molecular Subtypes of Breast Cancer Using Machine Learning Models Based on Multimodal MRI
Mengying XU ; Pan ZHANG ; Chunhua LI ; Jian LI ; Zihan HONG ; Bing CHEN
Chinese Journal of Medical Imaging 2025;33(10):1043-1048,1055
Purpose To explore the value of machine learning models based on synthetic MRI,dynamic contrast-enhanced MRI(DCE-MRI)and diffusion weighted imaging(DWI)parameters in identifying molecular subtypes of breast cancer.Materials and Methods A retrospective analysis was conducted on the data of 292 patients who underwent synthetic MRI,DCE-MRI and DWI examinations from September 2020 to September 2024 in Ningxia Medical University General Hospital before surgery and were pathologically confirmed to have breast cancer postoperatively.Patients were randomly divided into training and test sets using a ratio of 7:3.Multiple parameters were obtained from the synthetic MRI,DCE-MRI and DWI images.Variance analysis were used to screen the characteristic parameters among molecular subtype groups.Five machine learning models were established based on the selected characteristic parameters,and receiver operating characteristic curves were plotted to calculate the area under the curve among the molecular subtype groups.Results The support vector machine model exhibited the highest overall performance,with an area under the curve of 0.972,accuracy of 82.5%,specificity of 94.76%and sensitivity of 82.14%in the test set.This model's area under the curve values for differentiating luminal A,luminal B,human epidermal growth factor receptor-2 overexpression,and triple-negative groups in the training set were 0.979,0.925,0.971 and 0.982,respectively;in the test set,the area under the curve values were 0.973,0.873,0.956 and 0.955,respectively.Conclusion Machine learning models based on multimodal MRI parameters can assist clinicians in preoperatively determining the molecular subtypes of breast cancer and the support vector machine model shows relatively high comprehensive performance.
9.Effect of"Astragalus-Chinese yam"on cancer-induced fatigue in mice based on network pharmacology and behavior
Chen LIU ; Shiyi LI ; Xiaoxiao YIN ; Zihan KONG ; Qinghua LIU ; Hongli LI ; Chonggao YIN
Acta Laboratorium Animalis Scientia Sinica 2025;33(3):374-387
Objective To investigate the mechanism of action of the Astragalus-Chinese yam combination in treating cancer-related fatigue(CRF)in mice.Methods The active components and related targets of Astragalus-Chinese yam were obtained from the traditional Chinese medicine systems pharmacology database and analysis platform.CRF-associated targets were identified using the GeneCards database.Intersecting targets were analyzed using the DAVID database for gene ontology and kyoto encyclopedia of genes and genomes enrichment analyses.A network diagram depicting"drug-active component-intersecting targets-disease"was constructed using Cytoscape software,and a protein-protein interaction network was created to identify the top five core target proteins based on degree values.Molecular docking simulations were performed using Autodock Vina software.Twenty-five mice were divided randomly into a blank group and a modeling group in a 1∶4 ratio.After successfully establishing the CRF model using Lewis lung cancer cells,mice in the modeling group were further divided into model,Chinese yam(0.2 g/kg),Astragalus(0.6 g/kg),and Astragalus-Chinese yam combination groups(0.3+0.1 g/kg)(n=5 mice per group).The treatments were administered by gavage twice daily for 14 consecutive days.Grip-strength and forced-swimming tests were conducted.The mice were then euthanized and tissues were collected.The gastrocnemius muscles were weighed and stained with hematoxylin and eosin to reveal the muscle fiber morphology.Results A total of 23 effective active components of Astragalus-Chinese yam were identified through network pharmacology analysis,with 199 intersecting drug-disease targets.These targets mainly participated in biological processes such as protein phosphorylation through cellular components(cytoplasm,membrane,nucleus)and performed molecular functions such as protein binding.A total of 155 signaling pathways,including pathway in cancer and the phosphatidylinositol 3-kinase-protein kinase B signaling pathway,were involved in CRF.The critical targets of Astragalus-Chinese yam for CRF included serine/threonine kinase,tumor necrosis factor,epidermal growth factor receptor,B-cell lymphoma 2,and caspase 3.The active components quercetin and diosgenin interacted with the highest number of targets and demonstrated binding energies<-5.0 kJ/mol with the five core targets,indicating strong ligand-receptor binding affinity.Mice in the Chinese yam and Astragalus groups exhibited increased grip strength and prolonged swimming times compared with the model group.Gastrocnemius muscle volume and mass were increased,with well-organized muscle fibers and clear boundaries,and the effects were even more pronounced in the Astragalus-Chinese yam combination group.Conclusions Astragalus-Chinese yam treats CRF via a multi-target,multi-pathway approach,enhancing muscle strength and endurance in mice,improving gastrocnemius muscle volume and mass,and alleviating muscle atrophy,thereby mitigating the associated symptoms of CRF in mice.
10.Research on DIP Disease Cost Accounting Based on Parameter Allocation Method in County Public Hospitals
Wenli LIU ; Zihan LIN ; Jia LIU ; Yuxia LI ; Jiao ZHOU ; Jie PAN
Chinese Health Economics 2025;44(3):73-76
After the full implementation of Diagnosis-Intervention Packet(DIP)payment,driven by the dual mode of public welfare and economy,disease cost accounting,as a key link,promotes the transformation of hospitals from extensive income-driven development mode to refined operation management.Through literature analysis,it compares the characteristics,advantages and limitations of various methods of disease cost accounting.Combined with the current situation of case hospital cost accounting and the dilemma of disease cost accounting,the parameter distribution method combined with the service unit superposition method is used to realize the medical business cost and medical total cost accounting of the DIP disease in the case hospital.In addition,how to better implement DIP disease cost accounting in county-level public hospitals is discussed and suggested,in order to further improve the same type of hospitals and speed up the application of disease cost accounting results.


Result Analysis
Print
Save
E-mail