Saved in:
| Main Authors: | , |
|---|---|
| Format: | Artículo Open Access |
| Published: |
Wiley
2026
|
| Subjects: | |
| Online Access: | https://onlinelibrary.wiley.com/doi/10.1111/jerd.70100 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- Artificial Intelligence in the Laminate‐Veneer Workflow: A Systematic Review and Meta‐Analysis of Accuracy, Efficiency, and Esthetic Predictability Osama Hajeer Amal Hasan Journal of Esthetic and Restorative Dentistry ABSTRACT Background Artificial intelligence (AI) technologies are increasingly incorporated into restorative and esthetic dentistry; however, their reliability across the complete laminate‐veneer workflow remains uncertain. Objectives To systematically evaluate the accuracy, efficiency, and esthetic predictability of AI‐based systems applied to color prediction, margin‐line detection, computer‐aided design (CAD) automation, and esthetic simulation in laminate‐veneer procedures. Materials and Methods Five electronic databases (PubMed, Scopus, Web of Science, IEEE Xplore, and Google Scholar) were searched from inception to May 2025 in accordance with PRISMA 2020 guidelines. In vitro, in silico, and clinical studies applying AI or machine‐learning models to any step of the laminate‐veneer workflow and reporting quantitative outcomes were included. Random‐effects meta‐analyses were performed for compatible outcomes. Risk of bias was assessed using PROBAST‐AI, QUADAS‐AI, and ROBINS‐I, and the certainty of evidence was graded using the GRADE framework. Results Ten studies met the inclusion criteria. AI‐based color‐prediction models demonstrated a pooled mean color deviation (ΔE₀₀) of 1.85 (95% CI, 1.70–2.00), remaining below the clinical perceptibility threshold, with shade‐classification accuracy reaching 97.9% (96.3%–99.0%). Automated margin‐line detection achieved a pooled Dice similarity coefficient of 0.947 (0.931–0.963) and a mean deviation of 72 μm (55–89 μm), values within accepted laboratory tolerance. AI‐assisted CAD systems significantly reduced design time (Hedges g = −2.10 [−2.90 to −1.30]; p < 0.001) without compromising morphological accuracy. Clinical and perceptual studies showed moderate‐to‐high concordance between AI‐predicted and actual esthetic outcomes, although patient preference varied. The certainty of evidence was moderate for margin‐line accuracy and design‐time efficiency and low for color and esthetic prediction. Conclusions AI applications enhance efficiency and maintain clinically acceptable accuracy throughout the laminate‐veneer workflow. Despite promising results, larger clinical datasets, external validation, and standardized reporting are required before routine clinical implementation. 10.1111/jerd.70100 http://onlinelibrary.wiley.com/termsAndConditions#vor