Harnessing AI in prosthodontics and implant dentistry: An umbrella review of systematic evidence

TL;DR

This umbrella review summarizes AI applications in prosthodontics and implant dentistry, showing promise in diagnostic and planning tasks but highlighting the need for better-quality studies.

Contribution

A comprehensive synthesis of systematic reviews on AI in prosthodontics and implant dentistry, evaluating clinical applications and model performance.

Findings

AI showed high accuracy in implant type identification on radiographs (∼95.6%) and moderate accuracy in predicting treatment outcomes (62.4%–80.5%).

CNNs outperformed traditional algorithms in image-based tasks, but long-term outcome predictions were limited by data and biological variability.

Most included reviews had low methodological quality due to lack of protocol registration and incomplete bias assessment.

Abstract

To synthesize evidence from systematic reviews on artificial intelligence (AI) applications in prosthodontics and implant dentistry, focusing on clinical applications, AI model performance, and quality of evidence.

A comprehensive search was conducted in PubMed (MEDLINE), Scopus, Web of Science, Embase, and The Cochrane Library databases, identifying systematic reviews published from 2018 to 2025. Inclusion criteria comprised systematic reviews evaluating AI in prosthodontics or implant dentistry, published in English. Narrative reviews and reviews from other dental specialties were excluded. Two reviewers independently performed study selection and data extraction, and the methodological quality of the included systematic reviews was assessed using A Measurement Tool to Assess Systematic Reviews (AMSTAR‐2) tool. This umbrella review was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database (CRD420251067048).

Eleven systematic reviews were included. AI demonstrated substantial capability in prosthodontics for caries and fracture detection (with an accuracy of ∼82%–89%), automated tooth shade matching, and prosthesis design. In implant dentistry, AI algorithms accurately identified implant types on radiographs (∼95.6% pooled accuracy), optimized implant placement planning, and predicted treatment outcomes with moderate accuracy (62.4%–80.5%). Performance was strongest for radiographic identification and anatomic segmentation tasks in implant dentistry. It was more modest for preparation margin detection and objective shade matching in prosthodontics, as well as for multivariable prognosis and for detecting maxillary edentulous sites in implant dentistry. Convolutional neural networks (CNNs) consistently outperformed traditional algorithms in image‐based tasks. However, AI prediction of long‐term outcomes showed moderate performance due to data limitations and biological variability. Overall, although four reviews were rated as high quality, the majority exhibited low or critically low methodological quality, primarily due to a lack of a priori protocol registration and incomplete bias assessment.

AI applications in prosthodontics and implant dentistry may enhance diagnostic and planning workflows, especially for recognition and segmentation tasks. Nevertheless, most evidence comes from early‐stage, retrospective, or highly controlled studies, highlighting the need for prospective clinical validation and higher‐quality systematic reviews before routine clinical adoption can be recommended.

Figures