The Application of Deep Learning on CBCT in Dentistry

Area of Science:

• Dentistry
• Radiology
• Artificial Intelligence

Background:

• Cone beam computed tomography (CBCT) is crucial in dentistry for analyzing teeth and surrounding tissues.
• Manual CBCT analysis is time-consuming and user-dependent.
• Deep learning (DL) offers potential solutions to improve CBCT analysis efficiency and accuracy.

Purpose of the Study:

• To provide an overview of current deep learning applications in dental CBCT analysis.
• To highlight the potential and future research directions for DL in this field.

Main Methods:

• Systematic review of research articles from major scientific databases (PubMed, IEEE, Google Scholar, Web of Science) up to December 2022.
• Focus on studies applying deep learning models to CBCT images in dentistry.

Main Results:

• Deep learning models show significant progress in dental CBCT analysis, achieving clinician-level accuracy in many radiology image analysis tasks.
• Applications include automatic diagnosis, segmentation, classification of anatomical structures (teeth, nerves, bone, airway), and preoperative planning.
• Accuracy needs improvement in certain areas; ethical concerns and device variability pose challenges.

Conclusions:

• Deep learning combined with CBCT shows great potential to reduce dental image reading workload and serve as clinical decision-making aids.
• Further research is needed to address limitations and facilitate broader clinical implementation.