Related Experiment Video
Updated: Aug 5, 2026
07:32
Measuring Maxillary Posterior Tooth Movement: A Model Assessment using Palatal and Dental Superimposition
Published on: February 23, 2024
Prediction of Orthodontic Extraction Decisions Using Machine Learning Algorithms: A Retrospective Study
Alah Dawood Aldawoody1, Shehab Ahmed Hamad2
1Assistant professor of Orthodontics, Department of Pedodontics, Orthodontics and Preventive Dentistry, College of Dentistry, University of Mosul, Mosul, Iraq.
Journal of Clinical and Experimental Dentistry
|August 2, 2026
Summary
Machine learning models, particularly Random Forest, can objectively predict orthodontic extraction decisions. Key predictors include mandibular crowding and the IMPA angle, improving treatment consistency.
Area of Science:
- Orthodontics
- Machine Learning
- Dental Decision Support
Background:
- Orthodontic extraction decisions are subjective and challenging, especially in ambiguous cases.
- Clinical cues can be unclear, leading to inter-clinician variability.
Purpose of the Study:
- To develop machine learning (ML) models for predicting orthodontic extraction decisions.
- To identify key clinical predictors influencing these decisions.
Main Methods:
- Retrospective analysis of 200 patients (120 extraction, 80 non-extraction).
- Five ML models (LR, RF, SVM, DT, XGBoost) were trained and tested (70:30 ratio).
- Model performance evaluated using accuracy, sensitivity, specificity, and AUC-ROC; feature importance was calculated.
Main Results:
- The Random Forest (RF) model achieved the highest accuracy (93.5%) and AUC-ROC (0.95).
- XGBoost showed strong performance with 90.2% accuracy and 0.92 AUC-ROC.
- Mandibular crowding (0.28) and IMPA angle (0.22) were the most significant predictors.
Conclusions:
- Ensemble ML models, especially RF, offer an objective approach for orthodontic clinical decision support.
- These models can potentially reduce inter-clinician variation and enhance treatment planning consistency.
