Abstract
Background and Aim: Artificial Intelligence (AI) models, such as Convolutional Neural Networks (CNNs) and Artificial Neural Networks (ANNs), are increasingly be-ing utilized to analyze Cone Beam Computed Tomography (CBCT) images. This com-prehensive review aims to explore the diagnostic and prognostic accuracy of AI when applied to CBCT imaging in endodontic practices, highlighting its potential to enhance clinical outcomes.
Materials and Methods: A comprehensive electronic literature search was conducted using prominent electronic databases, including Google Scholar, PubMed, Medline, and Scopus. The search strategy spanned publications from 2000 to the present, employing targeted keywords such as “AI”, “machine learning”, “deep learning”, “endodontics”, “CBCT”, “advanced analytics”, “computational linguistics”, “automation”, “intelligent agents”, “probabilistic reasoning”, “CNN”, and “ANN".
Results: The initial search identified a total of 2072 articles, of which 37 met the stringent inclusion criteria for this review. The analysis revealed a significant concentration of research on key areas within endodontics, including the verification of working lengths, detection and projection of periapical pathologies, as-sessment of root canal morphologies, and predictions related to root fractures. Recent studies emphasize the growing focus on improving the precision and efficiency of endodontic diagnostics through AI-powered CBCT applications.
Conclusion: The findings of this review underscore the pivotal role that AI-enhanced CBCT imaging plays in advancing diagnostic accuracy and improving prognostic evaluations in endodontics. The integration of AI with CBCT heralds a new era of precision in endodontic care, facilitating more efficient workflows and standard-ized diagnostic practices while reducing the potential for human error.