Metabonomics Analysis of Brain Stem Tissue in Rats with Primary Brain Stem Injury Caused Death / 法医学杂志

OBJECTIVES@#To explore the potential biomarkers for the diagnosis of primary brain stem injury (PBSI) by using metabonomics method to observe the changes of metabolites in rats with PBSI caused death.@*METHODS@#PBSI, non-brain stem brain injury and decapitation rat models were established, and metabolic maps of brain stem were obtained by LC-MS metabonomics method and annotated to the HMDB database. Partial least square-discriminant analysis (PLS-DA) and random forest methods were used to screen potential biomarkers associated with PBSI diagnosis.@*RESULTS@#Eighty-six potential metabolic markers associated with PBSI were screened by PLS-DA. They were modeled and predicted by random forest algorithm with an accuracy rate of 83.3%. The 818 metabolic markers annotated to HMDB database were used for random forest modeling and prediction, and the accuracy rate was 88.9%. According to the importance in the identification of cause of death, the most important metabolic markers that were significantly up-regulated in PBSI group were HMDB0038126 (genipinic acid, GA), HMDB0013272 (N-lauroylglycine), HMDB0005199 [(R)-salsolinol] and HMDB0013645 (N,N-dimethylsphingosine).@*CONCLUSIONS@#GA, N-lauroylglycine, (R)-salsolinol and N,N-dimethylsphingosine are expected to be important metabolite indicators in the diagnosis of PBSI caused death, thus providing clues for forensic medicine practice.

Establishment and evaluation of intraoperative blood transfusion prediction model based on random forest algorithm / 中国输血杂志

【Objective】 To predict the risk factors of intraoperative blood transfusion by establishing a random forest algorithm prediction model, and to evaluate its prediction performance in clinical. 【Methods】 A total of 48 176 patients who underwent surgery from January 2014 to December 2017 in the First Medical Center of the Chinese People′s Liberation Army General Hospital were collected and divided into a blood transfusion group(n=5 035) and a non-transfusion group(n=43 141) according to whether blood was transfused or not during the operation, and the age, gender, weight, blood routine, coagulation test indicators, surgical grade, number of operations and anesthesia methods, and preoperative blood transfusion history between the two groups were compared and analyzed. All cases were randomly divided into training set(n=33 723) and the test set(n=14 453), using the sklearn function package in the computer programming language(Python V 3.9.0) to introduce the random forest algorithm, with 2 groups of different factors incorporated into the random forest algorithm to build the model, and the model was evaluated using the operating curve(ROC). 【Results】 1) There were statistically significant differences between the blood transfusion group and the non-transfusion group in terms of gender, age, blood routine, coagulation function, surgical grade, and preoperative blood transfusion history(P0.05); 3) In the established intraoperative blood model, the blood routine, coagulation function and general anesthesia had a great influence, with the cumulative importance > " 0.90" ; 4) The ROC analysis showed that the area under the ROC curve of the random forest model was 0.91 and 0.82 in the training set and the test set, which demonstrated a good predictive ability. 【Conclusion】 The intraoperative blood, using prediction model based on random forest method, can predict intraoperative blood use and blood transfusion risk factors.

Predicting passing rate for VMAT validation using machine learning based on plan complexity parameters / 中华放射医学与防护杂志

Objective:To establish a prediction model using the random forest (RF) and support vector machine (SVM) algorithms to achieve the numerical and classification predictions of the gamma passing rate (GPR) for volumetric arc intensity modulation (VMAT) validation.Methods:A total of 258 patients who received VMAT radiotherapy in the 1 st Affiliated Hospital of Wenzhou Medical University from April 2019 to August 2020 were retrospectively selected for patient-specific QA measurements, including 38 patients who received VMAT radiotherapy for head and neck, and 220 patients who received VMAT radiotherapy for chest and abdomen. Thirteen complexity parameters were extracted from the patient′s VMAT plans and the GPRs for VMAT validation under the analysis criteria of 3%/3 mm and 2%/2 mm were collected. The patients were randomly divided into a training cohort (70%) and a validation cohort (30%) , and the complexity parameters for the numerical and classification predictions were screened using the RF and minimum redundancy maximum correlation (mRMR) method, respectively. Complexity models and mixed models were established using PTV volume, subfield width, and smoothness factors based on the RF and SVM algorithms individually. The prediction performance of the established models was analyzed and compared. Results:For the validation cohort, the GPR numerical prediction errors of the complexity models based on RF and SVM under the two analysis criteria are as follows. The root-mean-square errors (RMSEs) under the analysis criterion of 3%/3 mm were 1.788% and 1.753%, respectively; the RMSEs under the analysis criterion of 2%/2 mm were 5.895% and 5.444%, respectively; the mean absolute errors (MAEs) under the analysis criterion of 3%/3 mm were 1.415% and 1.334%, respectively, and the MAEs under the analysis criteria of 2%/2 mm were 4.644% and 4.255%, respectively. For the validation cohort, the GPR numerical prediction errors of the mixed models based on RF and SVM under the two analysis criteria were as follows. The RMSEs under the analysis criterion of 3%/3 mm were 1.760% and 1.815%, respectively; the RMSEs under the analysis criterion of 2%/2 mm were 5.693% and 5.590%, respectively; the MAEs under the analysis criterion of 3%/3 mm were 1.386% and 1.319%, respectively, and the MAEs under the analysis criteria of 2%/2 mm were 4.523% and 4.310, respectively. For the validation cohort, the AUC result of the GPR classification prediction of the complexity models based on RF and SVM were 0.790 and 0.793, respectively under the analysis criterion of 3%/3 mm and were 0.763 and 0.754, respectively under the analysis criterion of 2%/2 mm. For the validation cohort, the AUC result of the GPR classification prediction of the mixed models based on RF and SVM were 0.806 and 0.859, respectively under the analysis criterion of 3%/3 mm and were 0.796 and 0.796, respectively under the analysis criterion of 2%/2 mm cohort.Conclusions:Complexity models and mixed models were developed based on the RF and SVM method. Both types of models allow for the numerical and classification predictions of the GPRs of VMAT radiotherapy plans under analysis criteria of 3%/3 mm and 2%/2 mm. The mixed models have higher prediction accuracy than the complexity models.

CT texture analysis for differential diagnosis of renal chromophobe cell carcinoma and renal oncocytoma / 中国医学影像技术

Objective: To investigate the value of texture analysis based on enhanced renal CT for identification of chromophobe cell renal carcinoma (CCRC) and renal oncocytoma (RO). Methods: CT images of 64 patients with CCRC and 31 with RO were retrospectively analyzed. ITK-SNAP version 4.11.0 software was used to delineate the region of interest, and A.K.Version v3.0.0.R software was used to extract texture features. Random forest model was established using texture features included in random forest algorithm. Logistic regression was used to evaluate the discriminative the efficacy of the established models for differential diagnosis of CCRC and RO. Results: The first 20 texture parameters selected with random forest algorithm from corticomedullary phase, nephrographic phase and both of them, with weight values from high to low, were evaluated with Logistic regression, and the AUC values were 0.876, 0.861 and 0.945, respectively. Conclusion: Texture analysis based on enhanced renal CT images has clinical value in differential diagnosis of CCRC and RO.

Prediction of dental caries in 12-year-old children using machine-learning algorithms / 대한구강보건학회지

OBJECTIVES: The decayed-missing-filled (DMFT) index is a representative oral health indicator. Prediction of DMFT index is an important basis for the development of public oral health care projects and strategies for caries prevention. In this study, we used data from the 2015 Korean children's oral health survey to predict DMFT index and caries risk groups using statistical techniques and four different machine-learning algorithms.METHODS: DMFT prediction models were constructed using multiple linear regression and four different machine-learning algorithms: decision tree regressor, decision tree classifier (DTC), random forest regressor, and random forest classifier (RFC). Thereafter, their accuracies were compared.RESULTS: For the DMFT predictive model, the prediction accuracy of multiple linear regression and RFC were 15.24% and 43.27%, respectively. The accuracy of DTC prediction was 2.84 times that of multiple linear regression. The important feature of the machine-learning model, which predicts DMFT index and the caries risk group, was the number of teeth with sealants.CONCLUSIONS: Using data from the 2015 Korean children's oral health survey, which is considered big data in the field of oral health survey in Korea, this study confirmed that machine-learning models are more useful than statistical models for predicting DMFT index and caries risk in 12-year-old children. Therefore, it is expected that the machine-learning model can be used to predict the DMFT score.

Prediction of 3-mouth glycemic control in type 2 diabetes mellitus based on machine learning algorithm / 中华疾病控制杂志

Objective To evaluate the efficiency of Logistic regression algorithm and random forest algorithm in prediction of blood glucose control in patients with type 2 diabetes mellitus (T2DM) after 3 months, and explore the influencing factors of blood glucose control. Methods The data was extracted from baseline survey and follow-up information of patients with T2DM in Shunyi and Tongzhou Districts. The patient’s 3-month glycosylated hemoglobin which was more than 6.5% was chosen as the outcome categorical variable. The random forest algorithm and Logistic algorithm were used to establish the prediction model. The predictive efficiency was evaluated with the area under receive operating characteristic curve (AUC) and accuracy rate. Results Factors affecting the patient’s glycemic control included baseline fasting plasma glucose(P<0.001), duration of disease(P<0.001), smoking(P=0.026), static activity time(P=0.006), body mass index(overweight P=0.002, obesity P=0.011), bracelet use(P=0.028), and diabetes diet(P=0.002).The Logistic regression prediction model had an AUC of 0.738, a sensitivity of 72.9%, a specificity of 68.1%, and an accuracy of 71.2%. The random forest model had an AUC of 0.756, a sensitivity of 74.5%, a specificity of 69.5%, and an accuracy of 72.8%. Conclusions The efficiency of random forest is better than Logistic regression model, which can be applied to the prediction of blood glucose control and assist the management of diabetic patients.

Prostate cancer prediction using the random forest algorithm that takes into account transrectal ultrasound findings, age, and serum levels of prostate-specific antigen / 亚洲男科学杂志(英文版)

The aim of this study is to evaluate the ability of the random forest algorithm that combines data on transrectal ultrasound findings, age, and serum levels of prostate-specific antigen to predict prostate carcinoma. Clinico-demographic data were analyzed for 941 patients with prostate diseases treated at our hospital, including age, serum prostate-specific antigen levels, transrectal ultrasound findings, and pathology diagnosis based on ultrasound-guided needle biopsy of the prostate. These data were compared between patients with and without prostate cancer using the Chi-square test, and then entered into the random forest model to predict diagnosis. Patients with and without prostate cancer differed significantly in age and serum prostate-specific antigen levels (P < 0.001), as well as in all transrectal ultrasound characteristics (P < 0.05) except uneven echo (P = 0.609). The random forest model based on age, prostate-specific antigen and ultrasound predicted prostate cancer with an accuracy of 83.10%, sensitivity of 65.64%, and specificity of 93.83%. Positive predictive value was 86.72%, and negative predictive value was 81.64%. By integrating age, prostate-specific antigen levels and transrectal ultrasound findings, the random forest algorithm shows better diagnostic performance for prostate cancer than either diagnostic indicator on its own. This algorithm may help improve diagnosis of the disease by identifying patients at high risk for biopsy.