AI Prognostic Models in Endodontics: Scoping Review and Clinical Gap Analysis for Retreatment vs Microsurgery...

Area of Science:

• Endodontic treatment outcome prediction
• Artificial intelligence in dentistry
• Clinical decision support systems

Background:

• Existing artificial intelligence (AI) prognostic models in endodontics predict outcomes but lack comparative analysis for treatment selection.
• Failed endodontic cases often present a choice between nonsurgical retreatment (NS-ReTx) and endodontic microsurgery (EMS), requiring comparative decision support.
• This study addresses the gap in AI-driven comparative decision-making for endodontic retreatment options.

Purpose of the Study:

• To conduct a scoping review of AI prognostic models in endodontics.
• To evaluate the alignment of these AI models with clinical decision domains for selecting between NS-ReTx and EMS.
• To assess the readiness of current AI models for comparative treatment decision support.

Main Methods:

• A dual-track scoping review following PRISMA-ScR guidelines.
• Searched major databases (PubMed, Embase, Web of Science, etc.) through February 2026.
• Track 1: Identified AI models for endodontic outcomes. Track 2: Synthesized clinical decision domains for NS-ReTx vs. EMS. Assessed risk of bias and reporting quality.

Main Results:

• Six AI prognostic studies were identified; 11 clinical studies informed decision domains.
• AI models utilized radiographic, structured-variable, or hybrid approaches, all with internal validation only.
• No models provided treatment-conditional predictions comparing NS-ReTx and EMS; key variables were inconsistently represented.

Conclusions:

• Current AI prognostic models in endodontics are in early stages, offering only single-pathway predictions.
• These models are not yet structured to provide comparative decision support for choosing between NS-ReTx and EMS.
• Further development is needed to enable AI-driven comparative analysis for complex endodontic retreatment decisions.