Related Experiment Video
Updated: Jun 27, 2026
Digital Hybrid Model Preparation for Virtual Planning of Reconstructive Dentoalveolar Surgical Procedures
Published on: August 5, 2021
Practitioners' Perceptions of Augmented and Virtual Reality in Endodontic Treatment Planning: A Pilot Study
Marcel Reymus1, Nils Werner1, Falk Schwendicke1
1Department of Conservative Dentistry, Periodontology and Digital Dentistry, LMU University Hospital, LMU Munich, Munich, Germany.
Aim:
To develop a streamlined software workflow for converting cone-beam computed tomography (CBCT) data into augmented reality (AR) and virtual reality (VR) formats, and to evaluate dental practitioners' perceptions of these advanced visualisation modalities compared to conventional CBCT analysis for endodontic treatment planning.
Methodology:
A custom add-on for Blender (an open-source 3D modelling software), named VirtualEndo, was developed to automate the processing of AI-segmented CBCT scans into AR- and VR-compatible formats. Thirty dentists evaluated two diagnostically challenging cases using four methods: conventional CBCT analysis, AI-driven 3D segmentation, AR visualisation on iPad and VR visualisation with Meta Quest 3. The presentation order was randomised, and participants completed a standardised training protocol. Perceptions regarding diagnostic confidence, usability, information extraction, and clinical relevance were assessed using 4-point Likert scales. Statistical analysis employed Friedman tests with post hoc Wilcoxon signed-rank tests and Bonferroni correction.
Results:
Significant differences were found across all evaluated categories (p < 0.05). Conventional CBCT was rated significantly inferior to Segmentation, VR, and AR for information extraction (canal course: W = 0.68; canal number: W = 0.56) and usability (W = 0.40). No significant differences existed between the three advanced visualisation methods. Segmentation was most frequently selected as clinically most relevant (40%), followed by VR (20%). For canal detection, differences were small (W = 0.10) with no method demonstrating clear superiority.
Conclusions:
Modern 3D visualisation technologies were perceived as significantly superior to conventional 2D CBCT slice analysis, primarily by presenting pre-integrated anatomical models that reduce the perceived cognitive burden of mental 3D reconstruction from 2D slices. Screen-based segmentation was favoured for pragmatic workflow integration, though immersive technologies showed promise if adoption barriers are addressed.
