Objective·To establish a nomogram for the differential diagnosis of early systemic lupus erythematosus(SLE)and other autoimmune diseases based on laboratory indications,and to evaluate its efficacy.Methods·A total of 535 SLE patients admitted to the First Hospital of Lanzhou University from January 2017 to December 2021 were selected as SLE group,and 535 patients with other autoimmune diseases during the same period were selected as control group.Basic information and laboratory test indicators of the SLE group and control group were collected and compared.The SLE group and control group were randomly assigned to the training set and the validation set at a ratio of 7∶3,respectively.LASSO regression method and multivariate Logistic regression were used to select the main risk factors of SLE.The nomogram for differential diagnosis of early SLE(SLE nomogram)was established according to the selected main risk factors.Bootstrap method was used to conduct internal repeated sampling for 1 000 times to calibrate the nomogram.The receiver operator characteristic curve(ROC curve)and decision curve analysis(DCA)were performed to evaluate the differential diagnosis ability and the value in clinical application of SLE nomogram,respectively.The"DynNom"package of R language was used to convert the nomogram into an electronic calculator,and its consistency with SLE nomogram was verified by data from 3 groups of patients.Results·LASSO regression and multivariate Logistic regression identified six major risk factors for SLE,including antinuclear antibody(ANA),anti-double-stranded DNA(anti-dsDNA)antibody,anti-ribonucleoprotein antibody/anti-Simth antibody(anti-nRNP/Sm),anti-ribosomal P protein(anti-P)antibody,anti-nucleosome antibody(ANuA)and urinary protein(PRO),which were used to construct the SLE nomogram.The calibration curve of the SLE nomogram had standard errors of 0.009 and 0.015 in the training set and validation set,respectively,and its area under the curve(AUC)was 0.889 and 0.869,respectively.The results of DCA showed that when the risk threshold of SLE nomogram was 0.15?0.95,the model achieved more net benefit.The prediction results of the electronic calculator showed that when ANA(titer 1∶100)was positive in SLE patient No.1,the prevalence was 0.166;when both ANA(titer 1∶100)and ANuA(titer 1∶100)were positive in patient No.2,the prevalence was 0.676;when all of PRO,ANA(titer 1∶100),ANuA(titer 1∶100)and anti-P antibody(titer 1∶100)were positive in patient No.3,the prevalence was 0.990,which was consistent with the differential diagnosis results of the SLE nomogram.Conclusion·The established SLE nomogram based on ANA,anti-dsDNA antibody,anti-nRNP/Sm,anti-P antibody,ANuA and PRO and its conversion into an electronic calculator can effectively distinguish early SLE from other autoimmune diseases,and have important clinical application value.

Objective To investigate the value of a nomogram model based on contrast-enhanced CT radiomics in predicting the histological type of thymoma.Methods A total of 154 patients(101 in low-risk group and 53 in high-risk group)with thymoma confirmed by pathology were retrospectively selected.The cases were randomly divided into training set(n=107)and validation set(n=47)at a ratio of 7∶3.The three-dimensional volume of interest(VOI)of the whole lesion on the image from the arterial phase of contrast-enhanced CT was manually delineated,and the radiomics features were extracted.Based on the selected radiomics features,the radiomics model was constructed and the model Radiomics score(Radscore)was calculated.Clinical risk factors were screened to construct a clinical model,and a nomogram model was constructed by fusing Radscore and clinical risk factors.The receiver operating characteristic(ROC)curve,area under the curve(AUC),accuracy,sensitivity and specificity were compared to analyze the predictive efficacy and difference of different models for high-risk and low-risk thymoma.The decision curve and calibration curve were drawn to evaluate the clinical value and fitting performance of the nomogram model.Results Eleven radiomics features were selected to construct the radiomics model,and five clinical risk factors[myasthenia gravis(MG),morphology,border,surrounding tissue invasion and CT value in arterial phase]were used to construct the clinical model.In the training set,the AUC of the nomogram model(0.88)was higher than that of the radiomics model(0.80)and the clinical model(0.79),and the difference was statistically significant(Z=2.233,2.713,P=0.026,0.007,respectively).In the validation set,the AUC of the nomogram model was higher than that of the radiomics and clinical models,but the difference was not statistically significant.The calibration curve showed that the nomogram model had good fitting performance,and the decision curve showed that the nomogram model had high clinical benefit.Conclusion The nomogram model based on contrast-enhanced CT can effectively predict high-risk and low-risk thymoma,which is helpful to guide clinicians to make relevant decisions.

Objective To observe the value of CT radiomics-based nomogram for predicting difference of Lauren types of gastric cancers between endoscopic biopsy and postoperative pathology.Methods Totally 126 patients with gastric cancer diagnosed by surgical pathology were retrospectively analyzed.The patients were divided into concordant group(n=77)and inconsistent group(n=49)according to the concordance between endoscopic biopsy and postoperative pathology results or not,also divided into training set and validation set at the ratio of 2∶1.Clinical predictors were screened,then a clinical prediction model was constructed.Radiomics features were extracted based on venous-phase CT images and screened using L1 regularization.Radiomics models were constructed using 3 machine learning(ML)algorithms,i.e.decision trees,random forests and logistic regression.The nomogram based on clinical and the best ML radiomics model was constructed,and the efficacy and clinical utility of the above models and nomogram for predicting inconsistency of Lauren types of gastric cancers between endoscopic biopsy and postoperative pathology were evaluated.Results Patients'age,platelet count,and arterial-phase CT values of tumors were all independent predictors of inconsistency between endoscopic biopsy and postoperative pathology of Lauren types of gastric cancer.CT radiomics model using random forests algorithm showed better predictive efficacy among 3 ML models,with the area under the curve(AUC)of 0.835 in training set and 0.724 in validation set,respectively.The AUC of clinical model,radiomics model and the nomogram in training set was 0.764,0.835 and 0.884,while was 0.760,0.724 and 0.841 in validation set,respectively.In both training set and validation set,the nomogram showed a good fit and considerable clinical utility.Conclusion CT radiomics-based nomogram had potential clinical application value for predicting inconsistency of Lauren types of gastric cancers between endoscopic biopsy and postoperative pathology.

Objective To predict the risk of in-hospital death in patients with chronic heart failure(CHF)complicated by lung infections using interpretable machine learning.Methods The clinical data of 1415 patients diagnosed with CHF complicated by lung infections were obtained from the MIMIC-IV database.According to the pathogen type,the patients were categorized into bacterial pneumonia and non-bacterial pneumonia groups,and their risks of in-hospital death were compared using Kaplan-Meier survival curves.Univariate analysis and LASSO regression were used to select the features for constructing LR,AdaBoost,XGBoost,and LightGBM models,and their performance was compared in terms of accuracy,precision,F1 value,and AUC.External validation of the models was performed using the data from eICU-CRD database.SHAP algorithm was applied for interpretive analysis of XGBoost model.Results Among the 4 constructed models,the XGBoost model showed the highest accuracy and F1 value for predicting the risk of in-hospital death in CHF patients with lung infections in the training set.In the external test set,the XGBoost model had an AUC of 0.691(95%CI:0.654-0.720)in bacterial pneumonia group and an AUC of 0.725(95%CI:0.577-0.782)in non-bacterial pneumonia group,and showed better predictive ability and stability than the other models.Conclusion The overall performance of the XGBoost model is superior to the other 3 models for predicting the risk of in-hospital death in CHF patients with lung infections.The SHAP algorithm provides a clear interpretation of the model to facilitate decision-making in clinical settings.

Objective To predict the risk of in-hospital death in patients with chronic heart failure(CHF)complicated by lung infections using interpretable machine learning.Methods The clinical data of 1415 patients diagnosed with CHF complicated by lung infections were obtained from the MIMIC-IV database.According to the pathogen type,the patients were categorized into bacterial pneumonia and non-bacterial pneumonia groups,and their risks of in-hospital death were compared using Kaplan-Meier survival curves.Univariate analysis and LASSO regression were used to select the features for constructing LR,AdaBoost,XGBoost,and LightGBM models,and their performance was compared in terms of accuracy,precision,F1 value,and AUC.External validation of the models was performed using the data from eICU-CRD database.SHAP algorithm was applied for interpretive analysis of XGBoost model.Results Among the 4 constructed models,the XGBoost model showed the highest accuracy and F1 value for predicting the risk of in-hospital death in CHF patients with lung infections in the training set.In the external test set,the XGBoost model had an AUC of 0.691(95%CI:0.654-0.720)in bacterial pneumonia group and an AUC of 0.725(95%CI:0.577-0.782)in non-bacterial pneumonia group,and showed better predictive ability and stability than the other models.Conclusion The overall performance of the XGBoost model is superior to the other 3 models for predicting the risk of in-hospital death in CHF patients with lung infections.The SHAP algorithm provides a clear interpretation of the model to facilitate decision-making in clinical settings.

Objective To observe the value of dual-energy CT(DECT)radiomics combined with clinical and CT features for predicting differentiation degree of gastric adenocarcinoma(GAC).Methods Totally 254 patients with GAC were prospectively analyzed and divided into high-grade group(low differentiation GAC,n=88)and low-grade group(middle-high differentiation GAC,n=166)according to pathological results.The patients were also divided into training set(n=203,including 70 high-grade and 133 low-grade GAC)and verification set(n=51,including 18 high-grade and 33 low-grade GAC)at the ratio of 8∶2.Radiomics features were extracted and screened based on venous phase single-level(40,70,100 and 140 keV)DECT,and a multi-energy radiomics model was constructed to predict GAC classification.Univariate analysis and multivariate logistic regression were used to analyze clinical and CT features as well as DECT parameters in training set to construct a clinic-CT model.Then a combined model was constructed through combining clinic-CT model with radiomics model.The predictive efficacy of the models were evaluated,and the calibration degree of the combined model was assessed.Results The area under the curve(AUC)of clinic-CT model,multi-energy radiomics model and combined model was 0.74,0.75 and 0.78 in training set,and 0.73,0.77 and 0.78 in verification set,respectively.Except for AUC of combined model was higher than that of clinic-CT model in training set(P＜0.05),no significant difference of AUC was found among models in training set nor verification set(all P＞0.05).The calibration degree of combined model was good in both training set and verification set(both P＞0.05).Conclusion DECT radiomics combined with clinical and CT features could effectively predict differentiation degree of GAC.

Objective To understand the research hotspots and research trends about convolutional neural networks in the field of oncology imaging diagnosis by analyzing the characteristics of published literature at home and abroad over the past decade. Methods The SCI-E database was used as the data source to retrieve literature about convolutional neural networks in the field of oncology imaging diagnosis published from 2012 to 2022. The distribution characteristics of countries, institutions, journals, co-cited authors, and keywords of the studies were analyzed by CiteSpace software. Results A total of 1088 papers were eventually included, and they were mostly from China, the United States, and India. A total of 39 papers were published by Sun Yat-sen University, the research institution with the highest number of publications. Radiology Nuclear Medicine Medical Imaging was the journal with the highest number of publications. A total of 25 high-frequency keywords and 15 burst keywords were obtained. The formation of 12 author co-citation clusters such as image segmentation and lung nodule, as well as 11 keyword clusters such as automatic segmentation and breast cancer, was observed. Conclusion Current research on convolutional neural networks for oncology imaging diagnosis focuses on oncology segmentation, lung-nodule recognition, assisted diagnosis of breast cancer, and other high-frequency oncology.

OBJECTIVES: Observational studies have suggested an association between preterm birth and stroke in late adulthood, but it remains unclear whether the association is causal. The purpose of this study was to evaluate the causal effects of gestational age on stroke and to determine the pathophysiological mechanisms underlying the causal associations. METHODS: Two-sample Mendelian randomization (MR) was performed to assess the causal effects of fetal gestational duration, early preterm birth (EPB), preterm birth, or postterm birth on stroke and its subtypes. Two-step Mendelian randomization (TSMR) and multivariable Mendelian randomization (MVMR) were additionally used to determine the role of common stroke risk factors, including cardiovascular diseases, hypertension, pulmonary impairment, inflammation, and metabolic diseases, in mediating the causal associations between gestational age and stroke and its subtypes. RESULTS: Genetically predicted EPB increased the risk of cardioembolic stroke (CES; odds ratio [OR], 1.115; 95% confidence interval [CI], 1.036 to 1.200; p=0.004) and large artery stroke (LAS; OR, 1.131; 95% CI, 1.031 to 1.241; p=0.009). The TSMR results showed that EPB was associated with a lower forced expiratory volume in the first second and forced vital capacity ratio (FEV1/FVC) (β=-0.020; 95% CI, -0.035 to -0.005; p=0.009), which increased the risk of CES and LAS. Further MVMR analysis showed that the associations between EPB and stroke disappeared after adjustment for FEV1/FVC. CONCLUSIONS Our data demonstrate that EPB is causally associated with an elevated risk of CES and LAS, and that pulmonary dysfunction mediates the causal impact of EPB on CES and LAS.

【Objective】 To study the effect of air inlet modes on the temperature variation, fluctuation, uniformity and coefficient of variation(CV), so as to evaluate the stability and uniformity of the temperature field in refrigerated warehouse for blood. 【Methods】 The temperature changes of blood refrigerated warehouse under different modes of air inlet during compressor operation were analyzed. The stability of the temperature field in the storehouse was evaluated by the changes, fluctuation, uniformity, CV and deviation of temperature at each testing point. 【Results】 The average value of temperature in the storehouse, adopting air inlet via straight blow, was (4.98±0.92)℃, while that of air inlet via micro-hole mode was(4.15±0.25)℃, with significant differences between each other(P<0.05). As to the CV of temperature, air inlet via straight blow was significantly different from that via micro hole(P<0.01). The fluctuation, uniformity and deviation of temperature created by straight blow and micro hole were 1.85±1.11 vs 0.49±0.38, 1.00±0.68 vs 0.47±0.37, and 0.61±0.45 vs 0.27±0.21, respectively, with significant differences between each other(P<0.01). 【Conclusion】 Compared with straight blow, the mean temperature created by micro hole was closer to the median value (4℃) of the temperature range, i. e.(4±2)℃, during blood storage. Otherwise, micro hole demonstrated a smaller CV of temperature, and superior performance in fluctuation, uniformity and deviation of the temperature at the testing points, which was conducive to ensure the stability of storehouse temperature field.

Pyrroloquinoline quinone (PQQ), an important redox enzyme cofactor, has many physiological and biochemical functions, and is widely used in food, medicine, health and agriculture industry. In this study, PQQ production by recombinant Gluconobacter oxydans was investigated. First, to reduce the by-product of acetic acid, the recombinant strain G. oxydans T1 was constructed, in which the pyruvate decarboxylase (GOX1081) was knocked out. Then the pqqABCDE gene cluster and tldD gene were fused under the control of endogenous constitutive promoter P0169, to generate the recombinant strain G. oxydans T2. Finally, the medium composition and fermentation conditions were optimized. The biomass of G. oxydans T1 and G. oxydans T2 were increased by 43.02% and 38.76% respectively, and the PQQ production was 4.82 and 20.5 times higher than that of the wild strain, respectively. Furthermore, the carbon sources and culture conditions of G. oxydans T2 were optimized, resulting in a final PQQ yield of (51.32±0.899 7 mg/L), 345.6 times higher than that of the wild strain. In all, the biomass of G. oxydans and the yield of PQQ can be effectively increased by genetic engineering.
Fermentation ; Gene Knockout Techniques ; Gluconobacter oxydans ; genetics ; metabolism ; Industrial Microbiology ; methods ; Multigene Family ; genetics ; Organisms, Genetically Modified ; PQQ Cofactor ; biosynthesis ; genetics ; Promoter Regions, Genetic ; genetics