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.Application Analysis of Rehmanniae Radix in Medical Cases of Qing Court
Yan JIN ; Tiegui NAN ; Yihan WANG ; Zhilai ZHAN
Chinese Journal of Experimental Traditional Medical Formulae 2026;32(3):232-238
To gain an in-depth understanding of the clinical application of Rehmanniae Radix during the Qing Dynasty and to clarify its specifications and corresponding therapeutic effects, this study took Rehmanniae Radix in the prescriptions documented in Research on Medical Cases of the Qing Imperial Court as the research subject. According to historical medical literature, a comprehensive investigation was conducted on the specifications, therapeutic efficacy, frequency of use, dosage, and seasonal patterns of Rehmanniae Radix employed by imperial physicians. The findings revealed that Rehmanniae Radix in the medical cases of the Qing court was primarily classified into three categories: Xiaoshengdi, Zhongshengdi, and Dashengdi. Xiaoshengdi was also referred to as Xishengdi or Cishengdi, all denoting dried Rehmanniae Radix. The term Xishengdi was inconsistently defined in the literature. It should refer to the slender variant of dried Rehmanniae Radix and was utilized as a specific specification in the medical cases of the Qing court. In contrast, the wild fresh roots of Rehmanniae Radix, described as "as slender as fingers", were commonly documented as fresh Rehmanniae Radix in these medical cases. There were variations in Rehmanniae Radix size and grading between historical and contemporary standards. Furthermore, therapeutic differences were observed among Rehmanniae Radix specifications in the medical cases of the Qing court. Xiaoshengdi and Zhongshengdi exhibited slightly stronger blood-cooling and heat-clearing effects while maintaining a non-cloying Yin-nourishing property. In contrast, Dashengdi demonstrated a greater emphasis on Yin supplementation with relatively milder heat-clearing activity. In the medical cases of the Qing court, the dosage of Rehmanniae Radix in different specifications was usually 11.2-18.7 g per dose, typically administered twice daily. Rehmanniae Radix in different specifications exhibits variations in efficacy, which can provide evidence-based insights for precise clinical application.
3.Application Analysis of Rehmanniae Radix in Medical Cases of Qing Court
Yan JIN ; Tiegui NAN ; Yihan WANG ; Zhilai ZHAN
Chinese Journal of Experimental Traditional Medical Formulae 2026;32(3):232-238
To gain an in-depth understanding of the clinical application of Rehmanniae Radix during the Qing Dynasty and to clarify its specifications and corresponding therapeutic effects, this study took Rehmanniae Radix in the prescriptions documented in Research on Medical Cases of the Qing Imperial Court as the research subject. According to historical medical literature, a comprehensive investigation was conducted on the specifications, therapeutic efficacy, frequency of use, dosage, and seasonal patterns of Rehmanniae Radix employed by imperial physicians. The findings revealed that Rehmanniae Radix in the medical cases of the Qing court was primarily classified into three categories: Xiaoshengdi, Zhongshengdi, and Dashengdi. Xiaoshengdi was also referred to as Xishengdi or Cishengdi, all denoting dried Rehmanniae Radix. The term Xishengdi was inconsistently defined in the literature. It should refer to the slender variant of dried Rehmanniae Radix and was utilized as a specific specification in the medical cases of the Qing court. In contrast, the wild fresh roots of Rehmanniae Radix, described as "as slender as fingers", were commonly documented as fresh Rehmanniae Radix in these medical cases. There were variations in Rehmanniae Radix size and grading between historical and contemporary standards. Furthermore, therapeutic differences were observed among Rehmanniae Radix specifications in the medical cases of the Qing court. Xiaoshengdi and Zhongshengdi exhibited slightly stronger blood-cooling and heat-clearing effects while maintaining a non-cloying Yin-nourishing property. In contrast, Dashengdi demonstrated a greater emphasis on Yin supplementation with relatively milder heat-clearing activity. In the medical cases of the Qing court, the dosage of Rehmanniae Radix in different specifications was usually 11.2-18.7 g per dose, typically administered twice daily. Rehmanniae Radix in different specifications exhibits variations in efficacy, which can provide evidence-based insights for precise clinical application.
4.Mechanism of Lijin manipulation regulating scar formation in skeletal muscle injury repair in rabbits
Kaiying LI ; Xiaoge WEI ; Fei SONG ; Nan YANG ; Zhenning ZHAO ; Yan WANG ; Jing MU ; Huisheng MA
Chinese Journal of Tissue Engineering Research 2025;29(8):1600-1608
BACKGROUND:Lijin manipulation can promote skeletal muscle repair and treat skeletal muscle injury.However,the formation of fibrosis and scar tissue hyperplasia are closely related to the quality of skeletal muscle repair.To study the regulatory effect of Lijin manipulation on the formation of fibrosis and scar tissue hyperplasia is helpful to explain the related mechanism of Lijin manipulation to improve the repair quality of skeletal muscle injury. OBJECTIVE:To explore the mechanism of Lijin manipulation to improve the repair quality of skeletal muscle injury in rabbits,thereby providing a scientific basis for clinical treatment. METHODS:Forty-five healthy adult Japanese large-ear white rabbits were randomly divided into blank group,model group and Lijin group,with 15 rats in each group.Gastrocnemius strike modeling was performed in both model group and Lijin group.The Lijin group began to intervene with tendon manipulation on the 3rd day after modeling,once a day,and 15 minutes at a time.Five animals in each group were killed on the 7th,14th and 21st days after modeling.The morphology and inflammatory cell count of gastrocnemius were observed by hematoxylin-eosin staining,the collagen fiber amount was observed by Masson staining,the expression of interleukin-6 and interleukin-10 in gastrocnemius was detected by ELISA.The protein and mRNA expressions of paired cassette gene 7,myogenic differentiation factor,myoblastogenin,alpha-actin,transforming growth factor beta 1,and type Ⅰ collagen were detected by western blot and RT-PCR,respectively,and the expression of type Ⅰ collagen protein was detected by immunohistochemistry. RESULTS AND CONCLUSION:Hematoxylin-eosin staining and Masson staining showed that compared with the model group,inflammatory cell infiltration and collagen fiber content decreased in the Lijin group(P<0.01),and the muscle fibers gradually healed.ELISA results showed that compared with the model group,the expression of interleukin-6 in the Lijin group continued to decrease(P<0.05),and the expression of interleukin-10 increased on the 7th day after modeling(P<0.05)and then showed a decreasing trend(P<0.05).Western blot and RT-PCR results showed that compared with the model group,the protein and mRNA expressions of paired cassette gene 7,myogenic differentiation factor,myoblastogenin in the Lijin group were significantly increased on the 14th day after modeling(P<0.05),but decreased on the 21st day(P<0.05);the protein and mRNA expressions of alpha-actin,transforming growth factor beta 1,and type Ⅰ collagen in the Lijin group were significantly decreased compared with those in the model group(P<0.05).Immunohistochemical results showed that the expression of type Ⅰ collagen in the Lijin group was significantly lower than that in the model group(P<0.05).To conclude,Lijin manipulation could improve the repair quality of skeletal muscle injury by inhibiting inflammation,promoting the proliferation and differentiation of muscle satellite cells,and reducing fibrosis.
5.Spermine Synthase : A Potential Prognostic Marker for Lower-Grade Gliomas
Chen LIU ; Hongqi LI ; Xiaolong HU ; Maohui YAN ; Zhiguang FU ; Hengheng ZHANG ; Yingjie WANG ; Nan DU
Journal of Korean Neurosurgical Society 2025;68(1):75-96
Objective:
: The objective of this study was to assess the relationship between spermine synthase (SMS) expression, tumor occurrence, and prognosis in lower-grade gliomas (LGGs).
Methods:
: A total of 523 LGG patients and 1152 normal brain tissues were included as controls. Mann-Whitney U test was performed to evaluate SMS expression in the LGG group. Functional annotation analysis was conducted to explore the biological processes associated with high SMS expression. Immune cell infiltration analysis was performed to examine the correlation between SMS expression and immune cell types. The association between SMS expression and clinical and pathological features was assessed using Spearman correlation analysis. In vitro experiments were conducted to investigate the effects of overexpressing or downregulating SMS on cell proliferation, apoptosis, migration, invasion, and key proteins in the protein kinase B (AKT)/epithelialmesenchymal transition signaling pathway.
Results:
: The study revealed a significant upregulation of SMS expression in LGGs compared to normal brain tissues. High SMS expression was associated with certain clinical and pathological features, including older age, astrocytoma, higher World Health Organization grade, poor disease-specific survival, disease progression, non-1p/19q codeletion, and wild-type isocitrate dehydrogenase. Cox regression analysis identified SMS as a risk factor for overall survival. Bioinformatics analysis showed enrichment of eosinophils, T cells, and macrophages in LGG samples, while proportions of dendritic (DC) cells, plasmacytoid DC (pDC) cells, and CD8+ T cells were decreased.
Conclusion
: High SMS expression in LGGs may promote tumor occurrence through cellular proliferation and modulation of immune cell infiltration. These findings suggest the prognostic value of SMS in predicting clinical outcomes for LGG patients.
6.Association of Genetically Predicted Obesity and Stool Frequency: Evidence From an Observational and Mendelian Randomization Study
Ke HAN ; Xiangyao WANG ; Shimin CHEN ; Xiaotong NIU ; Yan WANG ; Jingyuan XIANG ; Nan RU ; Miao LIU ; Ningli CHAI ; Enqiang LINGHU
Journal of Neurogastroenterology and Motility 2025;31(2):267-275
Background/Aims:
Obesity is associated with several gastrointestinal (GI) disorders and has been identified as a potential risk factor for various GI symptoms. Bowel frequency is an important indicator of bowel function. However, the causal link between obesity and gastrointestinal motility remains uncertain. This study aims to determine the causal effect of overall and central obesity on stool frequency.
Methods:
Four obesity-related anthropometric indicators–body mass index, body fat percentage, waist circumference (WC), and waist-tohip ratio (WHR)–were investigated. Individual-level baseline information from the UK Biobank was used to explore observational associations between obesity and stool frequency. Additionally, summary-level data from published genome-wide association studies were subjected to two-sample Mendelian randomization (MR) analyses to examine causal associations.
Results:
For all 4 indicators of obesity, higher levels of obesity were associated with more frequent bowel movements after adjusting for demographic characteristics, lifestyle, and dietary factors. After rigorous screening, 482 body mass index single nucleotide polymorphisms (SNPs), 7 body fat percentage SNPs, 48 WC SNPs, and 287 WHR SNPs were identified as instrument variables for MR analysis. The MR results were generally consistent with observational findings, proving that the associations observed in the overall obesity indicators were causal. For central obesity, the association between WHR and stool frequency remained consistent in both analysis phases, whereas WC showed a multidirectional association.
Conclusions
Obesity-related anthropometric indicators were causally associated with increased stool frequency in the overall and central obesity groups. Weight loss could be a potential approach to improve gastrointestinal regularity in individuals with obesity.
7.Construction of a smart elderly education system based on the AHP method under the Healthy China strategy
Yuan WANG ; Nan YAN ; Langhu ZENG ; Xiaodan SONG ; Qifeng BAI ; Rui JING ; Yinghan WANG
Journal of Shenyang Medical College 2025;27(5):454-460
Objective:To construct a smart elderly education system that in line with the current situation under the Healthy China strategy and explore the constraints on its institutional implementation,in order to provide a basis for scientifically evaluating the implementation effect of this system.Methods:A questionnaire survey was conducted on 34 experts in related fields.The analytic hierarchy process(AHP)was used to perform quantitative and qualitative analyses on the construction of the smart elderly education system,and the weights of the importance of each scheme were obtained.Results:The constructed model of the smart elderly education system included 5 first-level indicators and 15 second-level indicators.Calculations of the judgment matrix for the 5 first-level indicators showed that the maximum eigenvalue(λmax)=5.071 4,the random consistency index(RI)=1.12,the consistency index(CI)=0.017 8,and the consistency ratio(CR)=0.015 9.With CR<0.1,the result met the consistency requirement.For the second-level indicators under each first-level indicator,pairwise comparison judgment matrices were constructed.The CR values of all 5 judgment matrices were<0.1.In summary,all judgment matrices passed the consistency test and were applicable.Among the first-level indicators,the top three by weights were guarantee mechanisms,demand orientation,and implementation effectiveness.Among the second-level indicators,the top three by combined weights were funding support capacity,health literacy improvement rate,and health demand identification mechanism.Conclusions:The indicator evaluation system for the smart elderly education system constructed based on AHP includes core dimensions such as guarantee mechanisms and demand orientation,responding to the key issues of institutional implementation.The selected indicators are both objective and practical,and can serve as reference tools for actual evaluation.
8.CDK8/19 Enhances the Anti-tumor Efficacy of Gastric Cancer by Regulating PARP Inhibitor Sensitivity
Jun-Di WANG ; Wan-Chang LIU ; Jian-Song LIU ; Tian-Run LI ; Yan TIAN ; Dan-Tong SUN ; Ze-Nan FAN ; Xiao-Man LI ; Jia-Dong WANG
Chinese Journal of Biochemistry and Molecular Biology 2025;41(9):1280-1297
Gastric cancer remains one of the most prevalent and lethal malignancies of the digestive tract worldwide,underscoring the urgent need for more effective targeted therapeutic strategies.Poly(ADP-ri-bose)polymerase(PARP)inhibitors have demonstrated remarkable efficacy in tumors with homologous recombination repair(HRR)deficiency;however,their clinical application in gastric cancer remains limited.Clinical evidence suggests that patients harboring Helicobacter pylori infection in combination with HRR gene mutations exhibit a significantly elevated risk of developing gastric cancer,thereby supporting the potential benefit of PARP inhibition in this setting.In this study,a kinase inhibitor library was screened in combination with the PARP inhibitor olaparib in gastric cancer cells.And we identify the cy-clin-dependent kinase 8/19(CDK8/19)inhibitor Senexin A as a compound that synergistically enhances the cytotoxic effect of PARP inhibition(P<0.05).Phenotypic validation using CCK-8 and colony for-mation assays demonstrated that the combination treatment significantly suppressed cellular proliferation and clonogenic potential compared to either monotherapy(P<0.0001).Mechanistically,alkaline comet assays revealed a significant increase in DNA damage in the combination treatment group relative to either single-agent group(P<0.0001),suggesting that the synergistic effect results from the exacerbation of DNA damage via impaired DNA repair mechanisms.In addition,treatment with CDK8/19 inhibitors a-lone markedly increased the formation of γH2AX and 53BP1 foci in irradiated gastric cancer cells(P<0.0001),indicating inhibition of DNA damage repair pathways.Transcriptome sequencing further re-vealed that CDK8/19 inhibition impacts critical cellular pathways,including DNA repair,cell cycle reg-ulation,and RNA splicing.Co-immunoprecipitation assays confirmed that inhibition of CDK8/19 kinase activity significantly reduces the phosphorylation level of PARP1,suggesting a potential regulatory inter-action.Immunohistochemical analysis of tumor and adjacent non-tumor tissues from gastric cancer pa-tients demonstrated that CDK8 is significantly overexpressed in tumor tissues,supporting its potential as both a prognostic biomarker and a therapeutic target.Collectively,this study elucidates a mechanistic ba-sis by which CDK8/19 inhibition enhances the sensitivity of gastric cancer cells to PARP inhibitors.These findings provide a strong rationale for the combined use of CDK8/19 and PARP inhibitors as a tar-geted therapeutic strategy and offer promising translational implications for advancing personalized medi-cine in gastric cancer treatment.
9.Investigation and analysis of hepatitis E virus in rodents from selected regions of the Inner Mongolia Autonomous Region
Shanshan WANG ; Xiaoyan SI ; Mengmeng ZHANG ; Man LUO ; Zhen YAN ; Lili LI ; Xiaowei NAN ; Zhongbing ZHANG
Chinese Journal of Zoonoses 2025;41(9):925-931
This study investigated the prevalence of hepatitis E virus(HEV)in rodents within the Inner Mongolia Autonomous Region.In 2024,liver,spleen,kidney,and lung tissue samples were collected from rodents in 11 leagues and cities across the Inner Mongolia Autonomous Region,including Hohhot,Baotou,Hulunbuir,Xing'an League,Tongliao,Chifeng,Xilin Gol League,Ulan-qab,Ordos,Bayannur,and Wuhai.Nested PCR(RT-PCR)was used to detect the conserved regions of the HEV open reading frame 1(ORF1)gene.The RT-PCR-positive products were sequenced,and phylogenetic and homology analysis of the obtained sequences was performed.A total of 816 rodents were captured in this investigation,including 319 Rattus norvegicus(39.09%,319/816),206 Musmusculus(25.25%,206/816),and 140 Mongolian gerbils(17.16%,140/816).The HEV infection rate among rodents in the In-ner Mongolia region was 3.68%(30/816).Sequencing of the RT-PCR-positive results and analysis of the nucleotide sequences yielded 30 HEV-positive rodent samples.Phylogenetic analysis identified these sequences as belonging to the HEV-C1 genotype.The prevalence of HEV was observed in brown rats(Rattus norvegicus)in five leagues and cities within the Inner Mongolia region,includ-ing Xing'an League,Chifeng City,Hulunbuir City,Xilin Gol League,and Tongliao City,with infection rates of 16.67%,10.00%,5.98%,3.30%,and 2.50%,respectively.Brown rats,a species of house rats,frequently inhabit areas near human residences and have close interactions with humans and livestock.Studies have shown that multiple subtypes of HEV can cause zoonotic infections.Therefore,strengthening the monitoring of pathogens carried by rodents in residential environments and optimizing the prevention and control of rodent-borne diseases will be essential.Timely dissemination of relevant infectious disease knowledge to local communities will also be crucial,to decrease the risk of human infection.
10.Prediction model of knee osteoarthritis based on ultrasound score,MRI score,and serum TGF-β1 and Cat D levels
Zhili WANG ; Danfeng XU ; Nan LI ; Yan JIAO ; Ruisong SHANG
Journal of China Medical University 2025;54(9):802-807
Objective To construct a prediction model for the progression of knee osteoarthritis(KOA)based on ultrasound score,magnetic resonance imaging(MRI)score,and serum levels of transforming growth factor-β1(TGF-β1)and cathepsin D(Cat D).Methods Clinical data from 270 patients with KOA in Hengshui People's hospital from December 2022 to June 2024 were retrospec-tively analyzed.The patients were randomly divided into a modeling set(n=189)and validation set(n=81)at a ratio of 7∶3.The patients in the modeling set were categorized into mild-to-moderate and severe groups based on the degree of disease progression.Mul-tivariate logistic regression analysis was used to identify factors influencing KOA progression,and a prediction model was constructed using R software.Results Multivariate logistic regression analysis showed that body mass index,knee injury history,ultrasound score,WORMS score,TGF-β1,and Cat D were significant predictors of KOA progression(P<0.05).A nomogram-based prediction model was developed using these variables.The areas under the curve(AUC)of the nomograms for predicting disease progression in the modeling and validation sets were 0.889 and 0.860,respectively.The calibration curves showed that the predicted probability was in good agreement with the actual probability.Conclusion The prediction model developed in this study is effective in identifying patients at high-risk of KOA progression and may servce as a valuable tool for clinical assessment and decision making.

Result Analysis
Print
Save
E-mail