Automatic recognition of landmarks on digital dental models

Area of Science:

• Biomedical Engineering
• Dental Research
• Computational Imaging

Background:

• Quantitative assessment of dental and orthodontic treatment outcomes is crucial for progress.
• Manual landmarking of 3D dental models is time-consuming, costly, and hinders large-scale studies.
• Existing methods for analyzing dental models are laborious and prone to errors.

Purpose of the Study:

• To develop and present techniques and software for automated 3D dental scan mapping.
• To overcome the limitations of manual landmarking in dental research.
• To facilitate quantitative outcome assessment in orthodontics.

Main Methods:

• A multi-step automated process for 3D dental scans: orientation determination, tooth separation and identification, and landmark detection.
• Implementation of these techniques into specialized software.
• Focus on automating Modified Huddard Bodemham (MHB) landmarking for cleft lip/palate patients.

Main Results:

• Successful development of techniques and software for automated 3D dental scan analysis.
• Demonstration of the software's capability to map dental scans and identify landmarks.
• The process is extendable to various predetermined landmarks beyond MHB.

Conclusions:

• The developed software offers a significant advancement over traditional plaster models and caliper methods.
• Automated 3D scan analysis streamlines dental research and supports clinical assessments.
• This technology serves as a foundation for automating routine clinical evaluations, such as the Index of Orthodontic Treatment Need (IOTN).