Development of a Machine Learning-based Model for Methimazole Dosage Adjustment in Youth With Hyperthyroidism
Authors' conclusion
This study introduces the first data-driven tool to guide MMI dosing in pediatric hyperthyroidism, which can improve clinical efficiency.
Study context
- Population
- Youth aged ≤18 years with hyperthyroidism
- Exposure
- Machine learning models for predicting optimal MMI dose
- Outcome
- MAE between predicted and actual MMI dosages
Methodological detailsmulticenter, external validation
- Multicenter
- External validation
Main finding
As reported in the abstract
MAE of best model (XGBoost)
XGBoost best performance, internal validation mean absolute error (MAE)
1.72 mg
XGBoost best performance, external validation MAE
1.08 mg
Observational data: associations, not effects.
Further findings
Feature importance (SHAP analysis)
- Previous MMI dose, T3 and free T4 levels identified as key predictors
Original publication
Development of a Machine Learning-based Model for Methimazole Dosage Adjustment in Youth With Hyperthyroidism. J Clin Endocrinol Metab. Published 7 October 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →