Related Experiment Video
Updated: Jul 9, 2026
05:49
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
Classification of dental implants using supervised deep learning from dental radiographs: A scoping review
Tamires Santos-Melo1, Gabriela Costa da Silva1, Lizandra Esper Serrano1
1Department of Prosthodontics, Faculty of Odontology, Rio de Janeiro State University, Rio de Janeiro, Brazil.
Imaging Science in Dentistry
|July 8, 2026
Summary
Supervised deep learning models show promise for classifying dental implants in radiographs. However, current methods relying on implant brands may limit generalizability, necessitating a focus on intrinsic radiographic features for future applications.
Area of Science:
- Artificial Intelligence in Dentistry
- Radiographic Image Analysis
- Machine Learning Applications
Background:
- Dental implant classification from radiographic images is crucial for treatment planning and monitoring.
- Supervised deep learning (DL) offers advanced capabilities for image analysis in dentistry.
- Existing reviews lack a comprehensive overview of DL applications for dental implant classification.
Purpose of the Study:
- To conduct a scoping review of evidence on supervised deep learning models for dental implant classification using radiographic images.
- To identify current methodologies, datasets, and limitations in this field.
Main Methods:
- A systematic search was performed across multiple databases (PubMed, Google Scholar, etc.).
- Studies evaluating supervised DL models on panoramic or periapical radiographs for implant classification were included.
- Data extraction and descriptive statistics were employed to synthesize findings.
Main Results:
- Nine studies published between 2020-2024 met the inclusion criteria.
- Convolutional neural networks were the predominant DL models used.
- Datasets varied widely in size (355-156,965 radiographs) and included multiple implant brands.
Conclusions:
- Current DL approaches for dental implant classification predominantly use implant brand/manufacturer, potentially limiting generalizability.
- Future research should prioritize intrinsic radiographic characteristics (e.g., macrogeometry, prosthetic connections) for enhanced model applicability.
- Ethical considerations regarding data protection in AI applications require further attention.
