ABSTRACT
Objective:To analyze the global research hotspots in the field of pediatrics based on the Essential Science Indicators (ESI) database and explore the inspiration to domestic editors and pediatrics researchers.Methods:The journal distribution, country (region) distribution, cooperation, organization distribution, funding, publication language, hot topic words and other data of highly cited papers in the field of pediatrics in ESI database were collected and analyzed.Results:A total of 682 highly cited pediatrics papers were collected from 77 pediatrics journals included in Science Citation Index(SCI). Most of the highly cited pediatrics papers (182) were found to be published in Pediatrics.All 682 paper were published in English and frequently, characterized by multiple authors, institutions and fund support.Of 682 highly cited pediatrics papers, 435 papers were published in the United States(the first), 123 papers in England(the second) and 86 paper in Canada(the third). Novel coronavirus pneumonia, coronavirus, SARS coronavirus, autism and multiple system inflammatory syndrome are the main frontiers of global pediatric research at present.Specifically, focal pediatric system diseases mainly include respiratory system diseases, digestive system diseases, cardiovascular diseases, etc. Conclusions:ESI-based analysis of global research hotspots in the field of pediatrics provides reference materials for domestic and foreign pediatrics researchers to understand the global academic frontiers and development trends in the field of pediatrics and select topics for future scientific research.More importantly, this analysis can help domestic editors of pediatrics journals to plan topics and organize hot papers, so as to improve the academic quality and international influence of the journals.
ABSTRACT
Objective:To explore the value of different machine learning models based on Gd-EOB-DTPA enhanced MRI hepatobiliary phase radiomics features in preoperative prediction of microvascular invasion (MVI) of hepatocellular carcinoma (HCC).Methods:The data of 132 patients with HCC confirmed by pathology in the First Affiliated Hospital of Soochow University from January 2015 to May 2020 were retrospectively analyzed, including 72 cases of positive MVI and 60 cases of negative MVI. According to the proportion of 7∶3, the cases were randomly divided into training set and validation set. The radiomics features of hepatobiliary phase images for HCC were extracted by PyRadiomics software. The clinical and radiomics features of the training set were screened by the least absolute shrinkage and selection operator (LASSO) regression with 5 fold cross-validation, and then the optimal feature subset was obtained. Six machine learning algorithms, including decision tree, extreme gradient boosting, random forest, support vector machine (SVM), generalized linear model (GLM) and neural network, were used to build the prediction models, and the ROC curves were used to evaluate the prediction ability of the models. DeLong test was used to compare the differences of area under the curve (AUC) for 6 machine learning algorithms.Results:Totally 14 features selected by LASSO regression were obtained to form the optimal feature subset, including 2 clinical features (maximum tumor diameter and alpha-fetoprotein) and 12 radiomics features. The AUCs of decision tree, extreme gradient boosting, random forest, SVM, GLM and neural network based on the optimal feature subset were 0.969, 1.000, 1.000, 0.991, 0.966, 1.000 in the training set and 0.781, 0.890, 0.920, 0.806, 0.684, 0.703 in the validation set, respectively. There were significant differences in the AUCs between extreme gradient boosting and GLM or neural network ( Z=2.857, 3.220, P=0.004, 0.001). The differences in AUCs between random forest and SVM, GLM, or neural network were significant ( Z=2.371, 3.190, 3.967, P=0.018, 0.001,<0.001). The difference in AUCs between SVM and GLM was statistically significant ( Z=2.621 , P=0.009). There were no significant differences in the AUCs among the other machine learning models ( P>0.05). Conclusion:Machine learning models based on Gd-EOB-DTPA enhanced MRI hepatobiliary phase radiomics features can be used to preoperatively predict MVI of HCC, particularly the extreme gradient boosting and random forest models have high prediction efficiency.
ABSTRACT
Objective:To explore the value of gadolinium-ethoxybenzyl- diethylenetriamine pentaacetic acid (Gd-EOB-DTPA) enhanced MRI nomogram model for preoperative prediction of Ki-67 expression in hepatocellular carcinoma (HCC).Methods:Data of 85 patients of HCC confirmed by postoperative pathology, who underwent preoperative Gd-EOB-DTPA enhanced MRI between September 2016 and August 2019 in the First Affiliated Hospital of Soochow University were retrospectively evaluated. According to postoperative immunohistochemistry Ki-67 index, the 85 patients were divided into Ki-67 low expression group(Ki-67 index ≤10%, n=20) and Ki-67 high expression group (Ki-67 index >10%, n=65). Clinical data (hepatitis, cirrhosis, etc.), qualitative imaging parameters (tumor margin, capsule, etc.) were compared by χ 2 test and quantitative parameters [lesion-to-normal parenchyma ratio-arterial phase (LNR-AP), lesion-to-normal parenchyma ratio-portal phase (LNR-PP), lesion-to-normal parenchyma ratio-equilibrium phase (LNR-EP) and lesion-to-normal parenchyma ratio-hepatobiliary phase (LNR-HBP)] were compared by independent sample t test. The above statistically significant parameters were included in multivariate logistic regression to identify the independent predictors of Ki-67 high expression and then the nomogram model for predicting Ki-67 expression of HCC was established. Results:alpha-fetoprotein (AFP) tumor margin, arterial rim enhancement between the Ki-67 low expression group and the Ki-67 high expression group had significant differences (χ 2 were 8.196, 10.538 and 4.717, respectively, P<0.05). LNR-AP, LNR-PP, LNR-EP and LNR-HBP between the two groups had significant differences ( t were 2.929, 2.773, 2.890 and 3.437, respectively, P<0.05).The result of multivariate logistic regression revealed that AFP≥20 μg/L, non-smooth tumor margin and low LNR-HBP were the independent predictors of Ki-67 high expression (odds ratio were 4.090, 3.509 and 0.042, respectively, P<0.05).The Gd-EOB-DTPA enhanced MRI nomogram model for predicting Ki-67 expression of HCC was established successfully. The Area under the receiver operating characteristic curve of the nomogram was 0.837 and the corrected predictive curve fitted the ideal curve, which suggested the model had a good predictive efficiency. Conclusion:Gd-EOB-DTPA enhanced MRI nomogram model has great value in preoperative prediction of Ki-67 expression of HCC, which provided a personalized prediction method for Ki-67 expression in patient with HCC.
ABSTRACT
Objective:To investigate the value of pericoronary adipose tissue histogram parameters based on coronary CT angiography (CTA) images for the differentiation of acute coronary syndrome and stable coronary artery disease.Methods:The clinical data and CTA images of 93 patients with coronary CTA examination in Suzhou Kowloon Hospital from 2013 to 2018 were analyzed retrospectively. There were 39 patients with acute coronary syndrome (acute coronary syndrome group) and 54 patients with stable coronary artery disease (stable coronary artery disease group). A region of interest (ROI) was drawn around the stenosis of the coronary arteries, with CT attenuation ranging from-190 to -30 HU to exclude non-adipose tissue. The CT attenuation of ROI excluding non-adipose were measured and histogram analysis was performed. The obtained parameters included the mean value, median value and the 5th, 10th, 45th, 55th, 70th and 95th percentiles. The differences in histogram parameters between the two groups were compared, and then the value of each parameter in differentiating acute coronary syndrome and stable coronary artery disease was evaluated based on receiver operating characteristic (ROC) analysis. The stepwise regression of multivariate logistic regression analysis was used to identify the useful features and establish the final prediction model. The ROC curve of the final model was calculated and its value was analyzed.Results:The mean, median and the 5th, 10th, 45th, 55th,70th and 95th percentile differences between the acute coronary syndrome group and the stable coronary artery disease group were statistically significant (all P<0.05). The ROC curve for the median and the 95th percentile had the same area under curve (AUC) of 0.73. The sensitivity, specificity and AUC of the diagnostic model established by multiple logistic regression were 82.1%, 89.1% and 0.90 respectively. Conclusion:CT attenuation histogram of pericoronary adipose tissue is of high value in differentiating acute coronary syndrome from stable coronary artery disease.