Establishment of risk prediction model for postoperative liver injury after non-liver surgery based on different machine learning algorithms / 陆军军医大学学报

Objective To construct a machine learning prediction model for postoperative liver injury in patients with non-liver surgery based on preoperative and intraoperative medication indicators.Methods A case-control study was conducted on 315 patients with liver injury after non-liver surgery selected from the databases developed by 3 large general hospitals from January 2014 to September 2022.With the positive/negative ratio of 1 ∶3,928 cases in corresponding period with non-liver surgery and without liver injury were randomly matched as negative control cases.These 1243 patients were randomly divided into the modeling group(n=869)and the validation group(n=374)in a ratio of 7∶3 using the R language setting code.Preoperative clinical indicators(basic information,medical history,relevant scale score,surgical information and results of laboratory tests)and intraoperative medication were used to construct the prediction model for liver injury after non-liver surgery based on 4 machine learning algorithms,k-nearest neighbor(KNN),support vector machine linear(SVM),logic regression(LR)and extreme gradient boosting(XGBoost).In the validation group,receiver operating characteristic(ROC)curve,precision-recall curve(P-R),decision curve analysis(DCA)curve,Kappa value,sensitivity,specificity,Brier score,and F1 score were applied to evaluate the efficacy of model.Results The model established by 4 machine learning algorithms to predict postoperative liver injury after non-liver surgery was optimal using the XGBoost algorithm.The area under the receiver operating characteristic curve(AUROC)was 0.916(95%CI:0.883～0.949),area under the precision-recall curve(AUPRC)was 0.841,Brier score was 0.097,and sensitivity and specificity was 78.95%and 87.10%,respectively.Conclusion The postoperative liver injury prediction model for non-liver surgery based on the XGBoost algorithm has effective prediction for the occurrence of postoperative liver injury.

Construction and validation of a risk prediction model of intermediate cesarean section for primiparous women with failed vaginal delivery trial / 中华全科医师杂志

Objective:To construct and validate a prediction model for the risk of intermediate cesarean delivery for primiparous women with failed vaginal trial of labor.Methods:Clinical data of 6 128 pregnant women who gave birth in the Affiliated Hospital of Jining Medical College from January 2019 to December 2020 were collected. The puerpera was randomly divided into train set ( n=4 290) and validation set ( n=1 838). The factors influencing the conversion to cesarean section in primiparous women with failed vaginal trial of labor were analyzed with univariate and binary multivariate logistic regression, and a risk prediction model was established based on the influencing factors. The predictive power of the model was assessed by receiver operating characteristic (ROC) curve and the Hosmer-Lemeshow goodness-of-fit test in train set and validation set. Results:Among 6 128 pregnant women 1 042 cases failed in vaginal trial of labor and were transferred to cesarean section. Univariate analysis showed age, occupation, gestational weight gain, days of gestation, body temperature before delivery, fetal heart condition at delivery, fetal abdominal circumference, Bishop score, premature rupture of membranes, gestational illness, mode of induction of labor, labor analgesia, and fetal orientation were significantly associated with converting to cesarean delivery (all P<0.05). The multivariate binary logistic regression analysis showed that the age, gestational weight gain, body temperature, gestational co-morbidities, days of gestation, premature rupture of membranes, amniotic fluid contamination, induction of labor, and abnormal occipital position were independent risk factors for intermediate cesarean delivery ( OR=1.03-8.06, all P<0.05); while height, occupation, Bishop score, and labor analgesia were protective factors for intermediate cesarean delivery ( OR=0.17-0.96, all P<0.05). A risk prediction model was constructed based on the risk factors and protective factors. In train set, the area under the ROC curve(AUC) of the model was 0.902 (95% CI: 0.89-0.92, P<0.001), with the best cutoff value of 0.138, the sensitivity and specificity were 0.837 and 0.825, respectively; and the Hosmer-Lemeshow goodness-of-fit test showed P=0.192. In validation set the AUC of the model was 0.917 (95% CI: 0.90-0.93, P<0.001), and the sensitivity and specificity were 0.826 and 0.851, respectively; the total correct rate of the model was 87.21% (1 603/1 838). Conclusion:The risk prediction model of failed vaginal trial of labor in primiparous women for intermediate cesarean delivery constructed in this study has good clinical prediction efficacy and high correctness rate.

Establishment of artificial neural network model for predicting lymph node metastasis in patients with stage Ⅱ-Ⅲ gastric cancer / 中华胃肠外科杂志

Objective: To establish a neural network model for predicting lymph node metastasis in patients with stage II-III gastric cancer. Methods: Case inclusion criteria: (1) gastric adenocarcinoma diagnosed by pathology as stage II-III (the 8th edition of AJCC staging); (2) no distant metastasis of liver, lung and abdominal cavity in preoperative chest film, abdominal ultrasound and upper abdominal CT; (3) undergoing R0 resection. Case exclusion criteria: (1) receiving preoperative neoadjuvant chemotherapy or radiotherapy; (2) incomplete clinical data; (3) gastric stump cancer.Clinicopathological data of 1231 patients with stage II-III gastric cancer who underwent radical surgery at the Fujian Medical University Union Hospital from January 2010 to August 2014 were retrospectively analyzed. A total of 1035 patients with lymph node metastasis were confirmed after operation, and 196 patients had no lymph node metastasis. According to the postoperative pathologic staging. 416 patients (33.8%) were stage Ⅱ and 815 patients (66.2%) were stage III. Patients were randomly divided into training group (861/1231, 69.9%) and validation group (370/1231, 30.1%) to establish an artificial neural network model (N+-ANN) for the prediction of lymph node metastasis. Firstly, the Logistic univariate analysis method was used to retrospectively analyze the case samples of the training group, screen the variables affecting lymph node metastasis, determine the variable items of the input point of the artificial neural network, and then the multi-layer perceptron (MLP) to train N+-ANN. The input layer of N+-ANN was composed of the variables screened by Logistic univariate analysis. Artificial intelligence analyzed the status of lymph node metastasis according to the input data and compared it with the real value. The accuracy of the model was evaluated by drawing the receiver operating characteristic (ROC) curve and obtaining the area under the curve (AUC). The ability of N+-ANN was evaluated by sensitivity, specificity, positive predictive values, negative predictive values, and AUC values. Results: There were no significant differences in baseline data between the training group and validation group (all P>0.05). Univariate analysis of the training group showed that preoperative platelet to lymphocyte ratio (PLR), preoperative systemic immune inflammation index (SII), tumor size, clinical N (cN) stage were closely related to postoperative lymph node metastasis. The N+-ANN was constructed based on the above variables as the input layer variables. In the training group, the accuracy of N+-ANN for predicting postoperative lymph node metastasis was 88.4% (761/861), the sensitivity was 98.9% (717/725), the specificity was 32.4% (44/136), the positive predictive value was 88.6% (717/809), the negative predictive value was 84.6% (44/52), and the AUC value was 0.748 (95%CI: 0.717-0.776). In the validation group, N+-ANN had a prediction accuracy of 88.4% (327/370) with a sensitivity of 99.7% (309/310), specificity of 30.0% (18/60), positive predictive value of 88.0% (309/351), negative predictive value of 94.7% (18/19), and an AUC of 0.717 (95%CI:0.668-0.763). According to the individualized lymph node metastasis probability output by N+-ANN, the cut-off values of 0-50%, >50%-75%, >75%-90% and >90%-100% were applied and patients were divided into N0 group, N1 group, N2 group and N3 group. The overall prediction accuracy of N+-ANN for pN staging in the training group and the validation group was 53.7% and 54.1% respectively, while the overall prediction accuracy of cN staging for pN staging in the training group and the validation group was 30.1% and 33.2% respectively, indicating that N+-ANN had a better prediction than cN stage. Conclusions: The N+-ANN constructed in this study can accurately predict postoperative lymph node metastasis in patients with stage Ⅱ-Ⅲ gastric cancer. The N+-ANN based on individualized lymph node metastasis probability has better accurate prediction for pN staging as compared to cN staging.

Analysis of risk factors influencing the overall survival and establishment of nomogram predicting model in patients with rectal cancer at T1 and T2 stage / 安徽医科大学学报

Objective @#To explore the independent risk factors affecting the prognosis，and to construct a nomogram model predicting overall of patients with rectal cancer at T1 and T2 stage．@*Methods @#Retrospective analysis was made on the data of 353 patients diagnosed as rectal cancer，who received the radical rectal resection．The collect- ed data were as follows : age，body mass index (BMI) ，carcinoembryonic antigen ( CEA) ，tumor size，histological type，T stage，N stage，tumor location and number of lymph nodes detected，which were used to perform Kaplan- Meier curve and Log-rank test for univariate analysis and Cox regression for multivariate analysis．The nomogram model was established to predict the overall survival of patients. @*Results @#Age≥60 years，Mucinous adenocarcino- ma，poorly differentiation ，T2 stage ，lymph node metastasis ，BMI ≥25 kg / m2 ，CEA ≥5 μg / L and number of lymph nodes detected ＜12 were associated with overall survival of patients with rectal cancer at T1 and T2 stage (all P＜0. 05) ．Cox regression showed that age≥60 years，T2 stage，mucinous adenocarcinoma，lymph node me- tastasis，CEA≥5 μg / L，BMI ≥25 kg / m2 and lymph node detection number ＜12 were independent risk factors. Based on the above independent risk factors，the nomogram model was constructed，and the predicted curve was in good agreement with the actual survival curve ( C-index = 0. 779) ．@*Conclusion @#Age≥60 years，T2 stage，mucin- ous adenocarcinoma，lymph node metastasis，CEA≥5 μg / L，BMI≥25 kg / m2 and the number of lymph nodes de- tected ＜12 are independent risk factors ，and the nomogram established in this study can effectively predict the prognosis of patients with rectal cancer at T1 and T2 stage.

Establishment of a nomogram model predicting risk factors of postoperative complications after radical gastrectomy for gastric cancer / 上海交通大学学报（医学版）

Objective: To explore the risk factors of postoperative complications after radical gastrectomy + D2 lymphadenectomy and establish a predictive nomogram model. Methods: From July 2016 to June 2019, 1 705 patients who received radical gastrectomy + D2 lymphadenectomy in the Department of Gastrointestinal Surgery, Renji Hospital, Shanghai Jiao Tong University School of Medicine were collected. According to Clavien-Dindo grading system, the postoperative complications were graded, and the risk factors of postoperative complications ≥grade Ⅱ were analyzed by χ2 test. Multivariate Logistic regression was used to analyze the independent risk factors of postoperative complications ≥grade Ⅱ. According to the selected independent risk factors, the nomogram model was established. For verification, above patients were used as the training set, and 612 patients undergoing the same operation in this department from July to December 2019 were used as the validation set. Results: A total of 416 (24.4%) gastric cancer patients had postoperative complications. Multivariate Logistic regression analysis showed that male (OR=1.507, P=0.002), age ≥60 years old (OR=1.962, P=0.001), maximum diameter of tumor ≥5 cm (OR=1.456, P=0.002) and total gastrectomy (OR=1.313, P=0.026) were independent risk factors for postoperative complications ≥ grade Ⅱ. Based on these independent risk factors, the nomogram was established and presented good discrimination and predictive consistency in training set and validation set. Conclusion: The nomogram based on these four independent risk factors has a good predictive performance in predicting postoperative complications after radical gastrectomy for gastric cancer, and has a certain clinical application and reference value.

Establishment of a visualized of prediction model for postpyloric placement of spiral nasoenteral feeding tube in neurocritical care patients / 中国实用内科杂志

OBJECTIVE: To analyze the factors influencing postpyloric placement of spiral nasoenteral feeding tube(NET) in neurocritical care patients and establish a visualized prediction model. METHODS: Patients in Neurological Intensive Care Unit(NICU)who undertook postpyloric placement of NET after receiving prokinetics from Apr 2012 to Mar 2018 were included for retrospective analysis. The patients were divided into the success and failure group base on whether the tube tip entered into duodenum(or beyond)or not confirmed by bedside X-ray 24 hours later. The baseline data, APACHE Ⅱ score(acute physiology and chronic health evaluation Ⅱ), AGI grade(acute gastrointestinal injury), therapeutic measures and agents administered were recorded. Univariate and multivariate Logistic regression analysis was used to identify the potential factors affecting the postpyloric placement of NET. Based on those factors, a predicting model was established and visualized into an easy-to-use nomogram. RESULTS: A total of 241 patients including146 male and 95 female were enrolled for the study, with an median age of 58 years, median APACHEⅡscore of 20, median AGI of Ⅰ.The placement succeeded in 119(49.4%) of 241 patients. Logistic regression analysis demonstrated that APACHE Ⅱ score, sedatives and analgesics, vasopressors and AGI grade were among the influencing factors. A prediction model with a ROC-AUC of 0.8002 were established and visualized into a nomogram. CONCLUSION: APACHE Ⅱ score, sedatives and analgesics, vasopressors and AGI grade are the factors influencing success of postpyloric NET placement in neurocritical care patients, which incorporate a predicting model that can be visualized into a nomogram. The nomogram provided intensivists an easy-to-use decision support tool in NET placements.

XGboost model in predicting readmission of patients with ischemic stroke recurrence within 90 d / 中华神经医学杂志

Objective To explore the predictive efficacy of XGboost model in predicting risk of relapse and re-admission within 90 d in patients with ischemic stroke,and provide basis for early screening and prevention of high-risk population with ischemic stroke.Methods The clinical data of 6070 primary ischemic stroke patients admitted to our hospital from January 2007 to July 2017 were retrospectively collected.XGboost model and multivariate Logistic regression model were utilized to screen out the influencing factors of relapse and re-admission within 90 d in patients with ischemic stroke.A predictive model was set up.Receiver operating characteristic (ROC) curve was drawn and compared.Sensitivity,specificity and Youden index were calculated and compared to evaluate the prediction performance of XGboost model.Results During the observation period,a total of 520 patients with relapsed ischemic stroke were observed within a period of 90 d,and the incidence density was 8.57％.Multivariate Logistic regression analysis showed that length of first hospital stay,hypertension,pulmonary infection,neutrophil percentage,red blood cell distribution width (variable coefficient),and alkaline phosphatase level were independent influencing factors for re-hospitalization within 90 d of ischemic stroke,(OR=1.016,P=0.000,95％CI:1.008-1.025;OR=4.598,P=0.000,95％CI:3.717-5.687;OR=1.452,P=0.025,95％CI:1.048-2.012;OR=1.013,P=0.006,95％CI:1.004-1.022;OR=1.161,P=0.000,95％CI:1.090-1.237;OR=1.003,P=0.023,95％CI:1.000-1.005).Analysis of importance of risk factors for re-admission of ischemic stroke using XGboost model showed that the top 6 factors were hypertension,red blood cell distribution width,direct bilirubin,length of hospital stay,pulmonary infection,and alkaline phosphatase,and the corresponding importance scores were 32,20,19,18,15 and 14,respectively.ROC curve analysis results indicated that the area under the ROC for re-admission for XGboost model was 0.792 (95％CI:0.717-0.762),which was improved by 5％ as compared with that for multivariate Logistic regression model (0.739 [95％CI:0.764-0.818]).The sensitivity was 89.30％ and the Youden index was 0.444 for XGboost model,which were significantly higher than those for multivariate Logistic regression model (77.3％,0.405).Conclusions XGboost model is superior to multivariate Logistic regression model in predicting recurrence and re-admission of first ischemic stroke patients within 90 d.This model is suitable for prediction and early diagnosis of re-admission of ischemic stroke,which is of great clinical value.

Development and evaluation of a predicting model of dose volume histograms of parotid in NPC IMRT planning / 中华放射肿瘤学杂志

Objective To study the mathematical predicting model of parotid DVH for the NPC IMRT planning, and its accuracy with the analysis of medical data. Methods 50 NPC radiotherapy treatment plans with same beam setup were chosen as sample data set, then their parotid DVHs and distance of voxels in the parotid to the target volumes were calculated with self-developed program to form the distance to target histogram ( DTHs);principal component analysis was applied to DVHs and DTHs to acquire their principal components ( PCs) ,and then nonlinear multiple variable regression was used to model correlation between the DTHs' PCs, parotids volume, PTVs and the DVHs. Another 10 plans were chosen as test data set to evaluate the efficacy and accuracy of the final model by comparing the DVHs calculated from our model with those calculated from the TPS. Results Up to 97% information of DTHs and DVHs can be represented with 2 to 3 components, the average fitting error of sample data set was (0±3. 5)%;in the 10 test cases, the shapes of DVH curves calculated from predicting model was highly the same with those from the TPS, the average modeling error was (-0.7± 4. 4)%,the accuracy of prediction was up 95%. Conclusions Our developed model can be used as a quality evaluating tool to predict and assure the dose distribution in parotid of NPC radiotherapy treatment planning effectively and accurately.

Establishment of a assessment model for the clinical efficacy of neoadjuvant chemotherapy in nasopharyngeal carcinoma / 重庆医学

Objective To establish a model to predict the clinical response of neoadjuvant chemotherapy for nasopharyngeal car‐cinoma ,and provide basis for the individual treatment .Methods The clinical data of 63 cases of advanced nasopharyngeal carcinoma patients who have received neoadjuvant chemotherapy in the past 2 years were analyzed retrospectively .Univariate and multivariate analyses were performed using the Logistic analyses to identify efficacy factors .Results The response rate in nasopharyngeal tumor and lymph node metastasis were 39 .7% and 50 .8% ,respectively .Single factor analysis showed that patients with no distant metas‐tasis ,cranial nerve inviolated ,EBV negative and high expression of Ki67 were more sensitive to therapy .Logistic analysis showed that the influencing factors for the effect of the new chemotherapy include :distant metastasis ,cranial nerve inviolated and EBV . Thus ,the prediction model would be:Logit= -0 .470 -2 .863 × distant metastasis + 1 .328 × cranial nerve invasion+ 3 .639 × EBV ,its sensitivity ,specificity ,positive predictive value and negative predictive value were 79 .4% ,82 .8% ,84 .4% and 77 .4% . Conclusion The distant metastasis ,cranial nerve invasion and EBV infection were important predictive factors for neoadjuvant chemotherapy of nasopharyngeal carcinoma .This model could be used to predict the response of patients with nasopharyngeal carci‐noma .

The Role of Predictive Microbiology in Food Safety Risk Assessment / 微生物学通报

With the development of the food industry in China,it has been found that food safety is becoming the biggest issue in the food manufacture and logistics. Accurate and timely to establish a risk assessment method in produce market is the challenge for food safety system. Predictive microbiology is a core early warning technology in the food safety risk assessment. According to the microorganism predicting model,the pathogen and spoilage microorganism's growth in food can be fast judgment in advance. And it plays an important part in controlling the growth of pathogen and the spoilage microorganism in food. This paper summarized the predictive microbiology model's establishment and the present research situation,and discussed the present situation and application of predictive microbiology in food safety risk assessment. The future trend of predictive microbiology in food safety risk assessment was prospected as well.