Tooth Position Recognition Using Deep Learning and Motion Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional dental mirror systems struggle to accurately identify the position of teeth due to limited view and symmetry, requiring human expertise, and existing automated solutions are not effective in determining the actual tooth position.
Innovation Solution
A tooth-position recognition system that includes an electronic device with a camera and inertial measurement unit, using deep-learning models and multi-layer perceptron classifiers to generate probability values for tooth region identification, combining pixel position comparisons, user images, and motion tracking to determine the accurate tooth position.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a dental mirror with limited lens view is used, then the device structure is simple, but the ability to photograph teeth is limited and cannot capture all tooth areas
Solution Approach 1:
The patent divides the tooth identification task into multiple segments: capturing multiple images from different positions, extracting features from each image, and combining these features through deep learning models to determine the complete tooth position. This segmentation allows the simple mirror device to achieve accurate identification by processing information in manageable parts.
Solution Approach 2:
The patent transitions from two-dimensional image analysis to three-dimensional spatial understanding by incorporating depth information and multiple viewing angles. The system uses the relationship between pixel positions in multiple images and spatial coordinates to accurately determine tooth position, adding a dimensional aspect that resolves the limited view problem.
2Device complexity
If traditional dental mirror is used, then the device is simple, but human experience and expertise are required to determine tooth position
Solution Approach 1:
The system enables self-service by automatically determining tooth position without requiring dentist expertise. The deep learning models process captured images and autonomously identify tooth positions, allowing ordinary users to perform what traditionally required professional dental knowledge and experience.
Solution Approach 2:
The patent replaces the mechanical/reliant system (requiring physical manipulation and expert judgment) with an automated computational system. The deep learning models substitute for human expertise by automatically analyzing images and determining tooth positions through algorithmic processing rather than manual assessment.
3Measurement precision
If multiple images and deep learning models are used, then the tooth position identification accuracy is improved, but the system complexity increases
Solution Approach 1:
The patent merges multiple image processing results and deep learning model outputs into a unified tooth position determination. By combining features from multiple images and integrating predictions from different model components, the system achieves high accuracy while managing complexity through synthesis rather than simple accumulation of separate systems.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A tooth-position recognition system includes an electronic device and a calculation device. The electronic device includes a first camera. The first camera is configured to capture a plurality of tooth images. The calculation device includes a second camera and a processor. The second camera is configured to capture a user image. The processor is configured to receive the tooth images, compare the corresponding position of each pixel in each tooth image to generate a depth map, and input the tooth images, the depth map, and a plurality of first tooth-region identifiers into a tooth deep-learning model. The tooth deep-learning model outputs a plurality of deep-learning probability values that are the same in number as the first tooth-region identifiers. The processor inputs the user image and the plurality of second tooth-region identifiers into a user-image deep-learning model.