Machine Learning Model for Identifying ADHD Comorbidity in Children with Tic Disorders Using SHAP Interpretation

A retrospective study of 1,364 children with tic disorders (889 with TD alone, 475 with TD+ADHD comorbidity) developed a machine learning classification model to improve detection of ADHD in tic disorder patients, where ADHD symptoms are often masked. The model used 28 clinical features with stratified training/test validation and SHAP interpretability analysis to support diagnostic accuracy and treatment planning.