Interactive Tooth Separation from Dental Model Using Segmentation Field.

Area of Science:

• Computer-aided design (CAD)
• Medical imaging
• Orthodontics

Background:

• Accurate tooth segmentation is vital for orthodontic virtual treatment planning using CAD systems.
• Existing segmentation methods struggle to balance accuracy, speed, and simplicity, especially with complex dental models.
• Challenges include diverse tooth shapes, crowding, and varying model quality.

Purpose of the Study:

• To present a novel, efficient framework for tooth segmentation on dental models.
• To achieve accurate and fast tooth separation for improved virtual treatment planning.
• To overcome limitations of previous segmentation approaches.

Main Methods:

• Developed a framework utilizing a segmentation field solved via a linear system.
• Employed a discrete Laplace-Beltrami operator with Dirichlet boundary conditions.
• Detected cutting boundaries from contour lines sampled from the scalar field, focusing on concave regions.

Main Results:

• The proposed method demonstrates effective tooth partitioning by leveraging the sensitivity to concave seams.
• The algorithm is robust to low-quality dental models and varying degrees of tooth crowding.
• Experimental comparisons validate its effectiveness against existing methods like morphologic skeleton segmentation.

Conclusions:

• The novel framework provides an effective and efficient solution for tooth segmentation in dental CAD.
• It offers improved accuracy and robustness compared to existing methods, particularly for complex cases.
• This facilitates more precise virtual orthodontic treatment planning.