Assessing the Diagnostic Performance of ChatGPT-5.0 versus Machine Learning in Orthodontics: A Comparative Analysis...

Area of Science:

• Dentistry
• Artificial Intelligence
• Machine Learning

Background:

• Accurate orthodontic extraction decisions require evaluating clinical and cephalometric variables.
• Machine learning (ML) algorithms are increasingly used to aid clinical decision-making.

Purpose of the Study:

• To evaluate ChatGPT-5.0's performance in distinguishing orthodontic extraction decisions.
• To compare ChatGPT-5.0's performance against five supervised ML algorithms.

Main Methods:

• 520 orthodontic records were analyzed, with expert orthodontist consensus as the reference standard.
• ChatGPT-5.0 and ML models (XGBoost, random forest, SVM, logistic regression, MLP) were compared using 5-fold cross-validation.
• Performance metrics included accuracy, sensitivity, specificity, precision, F1-score, and balanced accuracy.

Main Results:

• XGBoost achieved the highest accuracy (78.08%), followed closely by ChatGPT-5.0 (75.77%).
• ChatGPT-5.0 demonstrated significantly higher accuracy than random forest, SVM, logistic regression, and MLP.
• ChatGPT-5.0 exhibited the highest sensitivity (76.68%), while XGBoost had the highest specificity (82.15%).

Conclusions:

• ChatGPT-5.0 performs comparably to traditional ML models in orthodontic extraction decision-making.
• ChatGPT-5.0's high sensitivity is a notable advantage for identifying extraction cases.
• AI tools like ChatGPT-5.0 show potential to support orthodontic treatment planning.