Objective To construct large medical model named by "Huaxi HongYi"and explore its application effectiveness in assisting medical record generation. Methods By the way of a full-chain medical large model construction paradigm of "data annotation - model training - scenario incubation", through strategies such as multimodal data fusion, domain adaptation training, and localization of hardware adaptation, "Huaxi HongYi" with 72 billion parameters was constructed. Combined with technologies such as speech recognition, knowledge graphs, and reinforcement learning, an application system for assisting in the generation of medical records was developed. Results Taking the assisted generation of discharge records as an example, in the pilot department, after using the application system, the average completion times of writing a medical records shortened (21 min vs. 5 min) with efficiency increased by 3.2 time, the accuracy rate of the model output reached 92.4%. Conclusion It is feasible for medical institutions to build independently controllable medical large models and incubate various applications based on these models, providing a reference pathway for artificial intelligence development in similar institutions.

Objective To construct a "disease-syndrome combination" mathematical representation model for pulmonary nodules based on oral microbiome data, utilizing a multimodal data algorithm framework centered on dynamic systems theory. Furthermore, to compare predictive models under various algorithmic frameworks and validate the efficacy of the optimal model in predicting the presence of pulmonary nodules. Methods A total of 213 subjects were prospectively enrolled from July 2022 to March 2023 at the Hospital of Chengdu University of Traditional Chinese Medicine, Sichuan Cancer Hospital, and the Chengdu Integrated Traditional Chinese and Western Medicine Hospital. This cohort included 173 patients with pulmonary nodules and 40 healthy subjects. A novel multimodal data algorithm framework centered on dynamic systems theory, termed VAEGANTF (Variational Auto Encoder-Generative Adversarial Network-Transformer), was proposed. Subsequently, based on a multi-dimensional integrated dataset of “clinical features-syndrome elements-microorganisms”, all subjects were divided into training (70%) and testing (30%) sets for model construction and efficacy testing, respectively. Using pulmonary nodules as dependent variables, and combining candidate markers such as clinical features, lesion location, disease nature, and microbial genera, the independent variables were screened based on variable importance ranking after identifying and addressing multicollinearity. Missing values were then imputed, and data were standardized. Eight machine learning algorithms were then employed to construct pulmonary nodule risk prediction models: random forest, least absolute shrinkage and selection operator (LASSO) regression, support vector machine, multilayer perceptron, eXtreme Gradient Boosting (XGBoost), VAE-ViT (Vision Transformer), GAN-ViT, and VAEGANTF. K-fold cross-validation was used for model parameter tuning and optimization. The efficacy of the eight predictive models was evaluated using confusion matrices and receiver operating characteristic (ROC) curves, and the optimal model was selected. Finally, goodness-of-fit testing and decision curve analysis (DCA) were performed to evaluate the optimal model. Results There were no statistically significant differences between the two groups in demographic characteristics such as age and sex. The 213 subjects were randomly divided into training and testing sets (7 : 3), and prediction models were constructed using the eight machine learning algorithms. After excluding potential problems such as multicollinearity, a total of 301 clinical feature information, syndrome elements, and microbial genera markers were included for model construction. The area under the curve (AUC) values of the random forest, LASSO regression, support vector machine, multilayer perceptron, and VAE-ViT models did not reach 0.85, indicating poor efficacy. The AUC values of the XGBoost, GAN-ViT, and VAEGANTF models all reached above 0.85, with the VAEGANTF model exhibiting the highest AUC value (AUC=0.923). Goodness-of-fit testing indicated good calibration ability of the VAEGANTF model, and decision curve analysis showed a high degree of clinical benefit. The nomogram results showed that age, sex, heart, lung, Qixu, blood stasis, dampness, Porphyromonas genus, Granulicatella genus, Neisseria genus, Haemophilus genus, and Actinobacillus genus could be used as predictors. Conclusion The “disease-syndrome combination” risk prediction model for pulmonary nodules based on the VAEGANTF algorithm framework, which incorporates multi-dimensional data features of “clinical features-syndrome elements-microorganisms”, demonstrates better performance compared to other machine learning algorithms and has certain reference value for early non-invasive diagnosis of pulmonary nodules.

Based on the theory of qi， blood and fluids， and taking into account of the pathogenesis evolution process from constraint to phlegm， stasis and then mass in pulmonary nodules， an attempt has been made to construct a three-dimensional differentiation system for pulmonary nodules from the dimensions of time and space. The temporal progression of the early， middle， and late stages of pulmonary nodules reflects the pathological changes from constraint to phlegm and then stasis in the metabolism disorders of qi， blood and fluid. The spatial structures such as size， density， and morphology of pulmonary nodules reflect the pathological states of the duration， severity， and primary and secondary conditions of qi， blood and fluid metabolism disorders. Based on the temporal progression， the therapeutic principles have been proposed， which are dispelling pathogenic factors and promoting the use of beneficial factors to interrupt the growth momentum in the early stage， removing turbidity and dispersing phlegm to reduce the degree of nodules in the middle stage， and dispersing nodulation and eliminating abnormalities in the late stage. Based on the spatial structures， the suggested therapeutic methods are using wind herbs， employing multiple approaches to treat phlegm， and promoting blood circulation to resolve stasis， so as to provide theoretical reference for the systematic diagnosis and treatment of pulmonary nodules in traditional Chinese medicine.

Objective To establish an animal model of lung adenoma in BALB/c mice based on dynamic characterization by micro-computed tomography(CT).Methods Eighty female SPF-grade BALB/c mice were divided randomly into four groups:model low dose group(1 mg/g urethane,iP,once),model medium dose group(1 mg/g urethane,ip,once a week,followed by 2 weeks),model high dose group(1 mg/g urethane,ip,once a week,followed by 4 weeks),and blank group(equal volume of saline).Growth of lung nodules in the mice was monitored regularly using Micro-CT.Three-dimensional images of the lungs were drawn using the Analyze 12.0 system,and lung tissues were taken for histopathological examination(hematoxylin and eosin).Results Lung nodules with round high-density shadows were observed at week 11 in all model groups compared with the findings in the blank group.The rate of nodule formation increased with increasing modeling weeks,with rates of nodule formation in the model high,medium,and low dose groups of 93.8%,93.8%,and 87.5%,respectively,at week 21.Most mice had two to four,followed by one,and one to two nodules,respectively.The average maximum diameter of the lung nodules in the low dose group was significantly higher than the diameters in the medium-and high-dose groups(P＜0.05),but there was no significant difference in lung nodule volume among the three groups.Regarding pathological type,hematoxylin and eosin staining revealed that the tumors in all the model groups were lung adenomas.Conclusions Lung adenomas were successfully induced in all urethane dose groups of mice and growth of the lung nodules could be characterized by micro-CT.The rate of nodule formation was highest in the medium dose group,which developed a moderate number of lung adenomas and provided a stable model,and was thus considered the most suitable model for the study of lung adenomas in mice.

[This corrects the article DOI: 10.1016/j.apsb.2023.04.002.].

Mevalonate metabolism plays an important role in regulating tumor growth and progression; however, its role in immune evasion and immune checkpoint modulation remains unclear. Here, we found that non-small cell lung cancer (NSCLC) patients with higher plasma mevalonate response better to anti-PD-(L)1 therapy, as indicated by prolonged progression-free survival and overall survival. Plasma mevalonate levels were positively correlated with programmed death ligand-1 (PD-L1) expression in tumor tissues. In NSCLC cell lines and patient-derived cells, supplementation of mevalonate significantly up-regulated the expression of PD-L1, whereas deprivation of mevalonate reduced PD-L1 expression. Mevalonate increased CD274 mRNA level but did not affect CD274 transcription. Further, we confirmed that mevalonate improved CD274 mRNA stability. Mevalonate promoted the affinity of the AU-rich element-binding protein HuR to the 3'-UTR regions of CD274 mRNA and thereby stabilized CD274 mRNA. By in vivo study, we further confirmed that mevalonate addition enhanced the anti-tumor effect of anti-PD-L1, increased the infiltration of CD8⁺ T cells, and improved cytotoxic function of T cells. Collectively, our findings discovered plasma mevalonate levels positively correlated with the therapeutic efficacy of anti-PD-(L)1 antibody, and provided the evidence that mevalonate supplementation could be an immunosensitizer in NSCLC.

ObjectiveTo explore the effect and mechanism of Sishenwan-containing serum on aerobic glycolysis in human colon cancer HCT116 cells. MethodCell counting kit-8 （CCK-8） was used to detect the cell viability of colon cancer HCT116 cells after treatment with Sishenwan-containing serum （2.5%， 5%， and 10%） for 24， 48， 72 h. The concentration of lactic acid， the content of intracellular glucose， and the activity of hexokinase （HK） and fructose-6-phosphate kinase （PFK） in the cell culture medium were detected by the micro-method. The content of glucose transporter 1 （GluT1） mRNA was detected by Real-time quantitative polymerase chain reaction （Real-time PCR）. The protein expression of GluT1 and methyltransferase-like 3 （MettL3） was detected by Western blot. The expression of GluT1 in cells was detected by immunofluorescence and the level of N6-methyladenosine （m⁶A） RNA methylation was detected by colorimetry. ResultCompared with the normal serum， 2.5%， 5%， and 10% Sishenwan-containing serum had no significant effect on the viability of HCT116 cells at 24 h， while 10% Sishenwan-containing serum showed a significant inhibitory effect on the viability of HCT116 cells at 48 h （P<0.05）. Hence， 10% Sishenwan-containing serum was used in subsequent experiments， and the intervention time was 48 h. Compared with the normal serum， 10% Sishenwan-containing serum could reduce lactate production （P<0.05）， down-regulate glucose uptake （P<0.05）， and blunt the activities of HK and PFK， the key rate-limiting enzymes of glycolysis （P<0.05）. Meanwhile， 10% Sishenwan-containing serum could decrease the expression of GluT1 protein （P<0.01） and mRNA （P<0.05） and reduce the proportion of cells expressing GluT1 （P<0.01）. Compared with the normal serum， Sishenwan-containing serum also decreased the protein content of MettL3 （P<0.05） and the methylation level of m⁶A RNA （P<0.01）. ConclusionSishenwan can inhibit glycolysis in colon cancer cells， and its inhibitory mechanism may be related to reducing MettL3 overexpression， inhibiting m⁶A RNA methylation， and down-regulating GluT1 and the activities of intracellular aerobic glycolysis-related enzymes such as HK and PFK.

Nucleos(t)ide analogues (NAs), which are widely used as the first-line anti-hepatitis B virus (HBV) drugs in clinical practice, can effectively inhibit the replication of HBV DNA, significantly slow down disease progression in chronic hepatitis B (CHB) patients, and reduce the development of end-stage liver diseases such as liver failure and liver cancer. However, for some CHB patients receiving first-line NAs for 48 weeks or longer, serum HBV DNA is still persistently or intermittently higher than the lower detection of limit of sensitive nucleic acid detection reagents. After discussion by the authors, low-level viremia (LLV) is defined as follows: persistent LLV refers to the condition in which CHB patients, who receive entecavir, tenofovir disoproxil fumarate, or tenofovir alafenamide fumarate for ≥48 weeks, test positive for HBV DNA by two consecutive detections with sensitive quantitative PCR, with an interval of 3-6 months, but have an HBV DNA level of ＜2000 IU/ml; intermittent LLV refers to the condition in which patients test positive for HBV DNA intermittently by at least three consecutive detections with sensitive quantitative PCR, with an interval of 3-6 months, but have an HBV DNA level of ＜2000 IU/ml. For the diagnosis of LLV, the issues of poor compliance and drug-resistant mutations should be excluded. LLV might be associated with the increased risk of progression to liver fibrosis or hepatocellular carcinoma in patients with liver cirrhosis under NA treatment, but there are still controversies over whether the original treatment regimen with NAs should be changed after the onset of LLV. This article summarizes the incidence rate of LLV under NA treatment and the influence of LLV on prognosis and analyzes the possible mechanisms of the osnet of LLV, so as to provide a reference for the management of LLV in patients treated with NAs.

Objective To determine the association between plasma concentrations of brain derived neurotrophic factor (BDNF), neuron-specific enolase (NSE), and S100β, and the occurrence of delirium in critically ill patients. Methods Totally 65 patients in Intensive Care Unit (ICU) of Wuxi People's Hospital of Nanjing Medical University between June 2015 and February 2016 were included in the present study. Delirium diagnosis was used by confusion assessment method for the ICU (CAM-ICU). Plasma BDNF, NSE, and S100β concentrations were determined on day 1(T1), 3(T3), and 10(T10) after ICU admission. The day of ICU admission was defined as T0. Results Compared with the plasma BDNF level on T1 (0.23±0.22) μg/L, the plasma BDNF level on T3 (0.59±0.34) μg/L and T10 (0.24±0.21) μg/L were higher, especially for that on T3 with a significant difference (F=21.58, P=0.018). Plasma NSE level on T3 (1.68±0.25) μg/L was significantly higher than that on T1 (1.22±0.32) μg/L (F=10.24, P=0.042). Compared with those without delirium, the delirious patients had lower BDNF, higher NSE and S100β on T1, T3 and T10, of which the difference of BDNF [T1: (0.23±0.22) μg/L vs. (1.02±0.24) μg/L, F=116.25,P<0.01; T3: (0.59±0.34) μg/L vs. (1.55±0.36) μg/L, F=82.39, P<0.01; T10: (0.24±0.21) μg/L vs. (1.09±0.55)μg/L, F=50.93, P=0.003, and NSE (T1: (1.22±0.32) μg/L vs. (0.47±0.23) μg/L, F=94.30, P<0.01;T3:(1.68±0.25) μg/L vs. (0.79±0.28) μg/L, F=78.63, P=0.017; T10: (0.98±0.37) μg/L vs. (0.51±0.22) μg/L, F=70.95, P=0.026) reached significant differences. Conclusions Plasma BDNF and NSE are closely related to the occurrence of delirium in critically ill patients, especially for BDNF. Clinical monitoring of plasma levels of BDNF can help to predict the outcome of brain function in critically ill patients.

Objective To explore the prognostic value of soluble triggering receptor expressed on myeloid cells-1(sTREM-1) in patients with ventilator-associated pneumonia (VAP).Methods A total of 103 VAP patients were enrolled from June 2013 to May 2015 in the ICU of Wuxi People's Hospital Affiliated to Nanjing Medical University.The demographics and clinical data were collected,while serum sTREM-1,procalcitonin (PCT),C-reactive protein(CRP),clinical pulmonary infection score(CPIS) and acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) were measured.Patients were divided into the death group and the survival group according to 28 d survival.The differences in demographics and clinical data were compared between groups.The values of sTREM-1,PCT,CPIS and APACHE Ⅱ for predicting 28 d death were evaluated by receiver operating curves(ROC).The surviving curve was drawn by Kaplan-Meier method.The possible prognostic factors were analyzed by univadate and logistic multivariate analysis.Results There were 76 patients in the survival group and 27 patients in the death group,and there was no difference in demographics between two groups(P＞0.05).The serum sTREM-1,PCT,CPIS and APACHE Ⅱ were higher in the death group[(89.50±18.45) pg/mL,(823.86±182.74) pg/ mL,(7.20±1.74) and (19.58±3.43)] than those in the survival group[(54.09±12.71) pg/mL,(579.81±193.45) pg/mL,(4.79±1.93) and (17.23±3.12),all P＜0.05].The areas under the ROC of sTREM-1,PCT,CPIS and APACHE Ⅱ for predicting 28 d death were 0.84±0.04(95％CI:0.75-0.92,P＜0.01),0.65±0.05(95％CI:0.55-0.74,P=0.49),0.67±0.06(95％CI:0.55-0.79,P＜0.01),0.79±0.04(95％CI:0.70-0.87,P=0.03),respectively.Patients were assigned into two groups by the best cutoffpoint of sTREM-l=75.00 pg/mL,and Kaplan-Meier survival analysis showed that 28 d survival rate in the low sTREM-1 group was significantly higher than that in the high sTREM-1 group (82.5％ vs.63.4％,x2=3.96,P＜0.05).Multivariate logistic regression analysis showed that both sTREM-1 (OR=1.08,95％CI:1.04-1.13,P＜0.01) andAPACHE Ⅱ (OR=1.39,95％CI:1.15-1.67,P＜0.01) were risk factors associated with 28 d death.Conclusions Early serum sTREM-1 can be used as a reliable predictor for the outcome of patients with VAP.