Insights into Predicting Tooth Extraction from Panoramic Dental Images: Artificial Intelligence vs. Dentists

Abstract

Tooth extraction is one of the most frequently performed medical procedures, yet determining whether a tooth should be extracted is not always straightforward. Using 26,956 single teeth images from 1,184 panoramic radiographs, we trained a ResNet50 network to classify teeth as either extraction-worthy or preservable. The best AI model achieved ROC-AUC of 0.901, significantly outperforming dentists’ average of 0.797. AI models could serve as decision support tools to monitor at-risk teeth and reduce errors in extraction indications.

Kunpeng Xie

Oral & Maxillofacial Surgeon & Researcher

Doctoral candidate at RWTH Aachen University, developing clinically grounded tools for image analysis, computer-aided surgery, and augmented reality navigation.