Back to all papers

Artificial Intelligence for Cone-Beam Computed Tomography in Endodontics: A PRISMA-Aligned Narrative Synthesis of Evidence From the United States and Europe (2021-2026).

June 13, 2026pubmed logopapers

Authors

Kachmar V

Affiliations (1)

  • International Dental Program, Tufts University School of Dental Medicine, Boston, USA.

Abstract

Cone-beam computed tomography (CBCT) is the reference three-dimensional imaging modality for endodontic diagnosis and treatment planning, but its interpretation is time-consuming and varies between observers. Artificial intelligence (AI) and convolutional neural networks in particular have been applied to CBCT to automate root canal segmentation, periapical lesion detection, and morphologic classification. Currently, there is no focused synthesis of evidence from the United States and Europe. This review is a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020-aligned narrative synthesis. PubMed/MEDLINE, Scopus, Web of Science, Google Scholar, and IEEE Xplore were searched for primary studies published between January 1, 2021, and April 30, 2026. Studies were included if they used AI with CBCT for endodontic applications and had a first or corresponding author from an institution in the United States or Europe, and quality was appraised against the Checklist for Artificial Intelligence in Medical Imaging (CLAIM) and Quality Assessment of Diagnostic Accuracy Studies tailored to Artificial Intelligence (QUADAS-AI). Of 612 records screened, 17 studies were included, originating mainly from four research groups (the University of Pennsylvania, KU Leuven, the Medical University of Graz, and Stony Brook University), with two further single-study contributions. Tooth, pulp, and canal segmentation was the most mature task (Dice 0.85-0.97), and periapical lesion detection reached early clinical validation (sensitivity 0.80-0.97; specificity 0.84-1.00). No eligible studies from the United States or Europe were found on vertical root fracture detection, AI-based working length determination, American Association of Endodontists case difficulty assessment, or treatment outcome prediction. CBCT-based endodontic AI in these regions is advancing unevenly: segmentation and lesion detection are approaching clinical use, while several clinically important tasks remain unaddressed and external validation across devices is still limited.

Topics

Journal ArticleReview

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.