top of page

Artificial Intelligence
in Endodontics

Interpretable AI for personalised endodontic prognosis and clinical decision-making

Research Focus:

  • Radiomics analysis of periapical lesion core and boundary regions

  • Integration of imaging features with patient- and tooth-level clinical data

  • Interpretable machine learning for root canal retreatment prognosis

  • Multi-centre validation and translation into clinically useful decision-support tools
     

Key Study: Predicting Root Canal Retreatment Outcome Using a Machine Learning and Radiomics Approach (International Dental Journal. 2027)
Abstract: This proof-of-concept study developed an interpretable machine-learning framework to predict 12-month outcomes after nonsurgical root canal retreatment. Using 200 matched successful and failed cases, the model integrated radiomics features from periapical radiographs with preoperative clinical variables. The best-performing combined model achieved a stacked AUC of 0.902, outperforming models based on radiomics or clinical variables alone. Radiomics features from the boundary of periapical lesions made the greatest contribution to prediction, while greater lesion heterogeneity was associated with a higher risk of retreatment failure.

01 / 04
An interpretable machine-learning workflow combining radiomics features from periapical radiographs with clinical variables to predict 12-month outcomes after nonsurgical root canal retreatment.

All Related Publications:

bottom of page