AUTHOR=Ma Xiao-Hui , Shu Liqi , Jia Xuan , Zhou Hai-Chun , Liu Ting-Ting , Liang Jia-Wei , Ding Yu-shuang , He Min , Shu Qiang 

TITLE=Machine Learning-Based CT Radiomics Method for Identifying the Stage of Wilms Tumor in Children

JOURNAL=Frontiers in Pediatrics

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/pediatrics/articles/10.3389/fped.2022.873035

DOI=10.3389/fped.2022.873035

ISSN=2296-2360

ABSTRACT=Purpose: To develop and validate a machine learning-based CT radiomics method for preoperatively predicting the stages (stage I and non-stage I) of Wilms tumor (WT) in pediatric patients.
Methods: 118 cases with WT underwent contrast-enhanced computed tomography (CT) scan between 2014 and 2021 in our center were analyzed retrospectively, and were divided into two groups, namely stage I and non-stage I disease. Patients were randomly divided into training cohort (n= 94) and test cohort (n= 24). A total of 1781 radiomic features from seven feature classes were extracted from preoperative portal venous–phase images of abdominal CT. The Synthetic Minority Over-Sampling Technique (SMOTE) was used to handle the imbalanced dataset, then t-test and The Least Absolute Shrinkage and Selection Operator (LASSO) regularization were utilized for feature selection. Support Vector Machine (SVM) was deployed using the selected informative features to develop the predicting model. The performance of the model was evaluated according to its accuracy, sensitivity, and specificity. The receiver operating characteristic curve (ROC) and the area under the ROC curve (AUC) were also arranged to assess the model performance.
Results: The SVM model was fitted incorporating 15 radiomics features obtained by t-test and LASSO with respect to WT staging in the training dataset and demonstrated favorable performance in the testing dataset. Cross-validated AUC on training dataset was 0.79 with 95 percent confidence interval (CI) of 0.773-0.815 and coefficient of variation of 3.76%, while AUC on the test dataset was 0.81, and accuracy, sensitivity, and specificity were 0.79, 0.87, and 0.69, respectively.
Conclusions: Machine learning model of SVM based on radiomic features extracted from CT images accurately predicted WT stage I and non-stage I disease in pediatric patients preoperatively, which provided a rapid, non-invasive way for investigation of WT stages.