• Search
  • Advanced Search

Search

  • HOME
  • Search
Original Article
Infection
Construction and validation of predictive models for intravenous immunoglobulin–resistant Kawasaki disease using an interpretable machine learning approach
Linfan Deng, Jian Zhao, Ting Wang, Bin Liu, Jun Jiang, Peng Jia, Dong Liu, Gang Li
Clin Exp Pediatr. 2024;67(8):405-414.   Published online July 23, 2024
Question: Is there a reliable model to predict intravenous immunoglobulin (IVIG)-resistant Kawasaki disease (KD)?
Finding: We constructed 5 machine learning models to predict IVIG-resistant KD. Extreme gradient boosting (XGBoost) model was superior to logistic, support vector machine, light gradient boosting machine and multiple layers perception models. The SHAP (SHapley Additive exPlanations) value interpreted the contribution of each feature in XGBoost model.
Meaning: XGBoost model showed the excellent performance to predict IVIG-resistant KD with explainable and visualizable machine learning algorithm.
Review Article
Neurology
Big data analysis and artificial intelligence in epilepsy – common data model analysis and machine learning-based seizure detection and forecasting
Yoon Gi Chung, Yonghoon Jeon, Sooyoung Yoo, Hunmin Kim, Hee Hwang
Clin Exp Pediatr. 2022;65(6):272-282.   Published online November 26, 2021
· Big data analysis, such as common data model and artificial intelligence, can solve relevant questions and improve clinical care.
· Recent deep learning studies achieved 0.887–0.996 areas under the receiver operating characteristic curve for automated interictal epileptiform discharge detection.
· Recent deep learning studies achieved 62.3%–99.0% accuracy for interictal-ictal classification in seizure detection and 75.0%– 87.8% sensitivity with a 0.06–0.21/hr false positive rate in seizure forecasting.