A Fully Automated Method for 3D Individual Tooth Identification and Segmentation in Dental CBCT

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Dental Technology

Background:

• Accurate segmentation of individual teeth in 3D cone-beam computerized tomography (CBCT) is difficult due to complex anatomical structures.
• Existing methods face challenges in separating teeth from adjacent teeth and alveolar bone.
• High-dimensional CBCT data and limited training datasets pose computational and learning challenges.

Purpose of the Study:

• To propose a fully automated method for identifying and segmenting 3D individual teeth from dental CBCT images.
• To address the challenges of tooth separation and computational complexity in 3D CBCT analysis.
• To provide an effective framework for digital dentistry applications.

Main Methods:

• A deep learning-based hierarchical multi-step model was developed.
• Automatic generation of 2D panoramic images from 3D CBCT data to reduce dimensionality.
• Identification of 2D individual teeth and capture of loose/tight regions of interest (ROIs) for 3D segmentation.

Main Results:

• The method achieved an F1-score of 93.35% for tooth identification.
• A Dice similarity coefficient of 94.79% was obtained for individual 3D tooth segmentation.
• Experimental results demonstrate high accuracy and effectiveness.

Conclusions:

• The proposed automated method offers an effective solution for 3D individual tooth segmentation from CBCT images.
• The hierarchical deep learning approach successfully overcomes segmentation challenges.
• This framework holds significant potential for clinical and practical applications in digital dentistry.