Automated 3D Tooth Segmentation With AI Boundary Curves
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional tooth segmentation methods in dental scans suffer from inaccuracies such as incomplete masking of tooth areas, invasion of masks into adjacent teeth or gums, and incorrect grouping of multiple teeth as one, requiring manual effort and skilled operators, and are limited by algorithmic constraints, especially when dealing with partial teeth data.
Innovation Solution
An automated method using artificial intelligence neural networks for tooth detection and boundary generation, including mesh parameterization, to accurately segment teeth by generating a tooth boundary curve, which converts 3D scan data to 2D images, extracts tooth data using bounding boxes and landmarks, and uses curvature data inversion for precise boundary definition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual tooth boundary designation is performed by operator, then flexibility and adaptability are maintained, but accuracy decreases and time consumption increases
Solution Approach 1:
The system enables automated tooth segmentation where the computer automatically identifies and segments teeth from scan data using deep learning algorithms, eliminating the need for manual operator intervention in boundary designation while maintaining high accuracy through AI-based detection
Solution Approach 2:
The patent replaces the manual mechanical process of operator-based boundary drawing with an automated computational system using deep learning neural networks that automatically detect tooth boundaries and generate segmentation masks from three-dimensional scan data
2Device complexity
If conventional masking method is used for tooth segmentation, then simplicity is maintained, but segmentation accuracy deteriorates with incomplete masking, invasion into adjacent teeth, and incorrect tooth grouping
Solution Approach 1:
The patent replaces conventional simple masking algorithms with deep learning-based automated segmentation that uses neural networks to accurately identify tooth boundaries, preventing mask invasion into adjacent teeth and gums while correctly separating individual teeth even in complex arrangements
Solution Approach 2:
The system transforms the segmentation approach by changing from fixed threshold-based masking to adaptive deep learning models that learn optimal segmentation parameters from training data, enabling accurate tooth boundary detection that adapts to varying tooth geometries and scan qualities
3Adaptability or versatility
If deep learning model for partial teeth is created, then segmentation capability for partial teeth is improved, but memory usage and processing speed deteriorate
Solution Approach 1:
The patent develops a universal deep learning model that can handle both complete and partial teeth segmentation using the same network architecture and training approach, eliminating the need for separate specialized models for different tooth completeness scenarios, thereby optimizing memory efficiency while maintaining versatility
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
An automated method includes detecting a tooth of the scan data using a first artificial intelligence neural network, extracting a tooth scan data from the scan data based on a result of a tooth detection, generating a tooth mapped data corresponding to a predetermined space based on the tooth scan data, generating the tooth boundary curve by inputting the tooth mapped data to a second artificial intelligence neural network and mapping the tooth boundary curve to the scan data.