A panorama constant-tooth identification method and device based on deep learning

A deep learning and panorama technology, applied in image data processing, instruments, calculations, etc., can solve problems such as operator eye fatigue, difficulty in manually identifying tooth positions, and reduce the accuracy and credibility of judgment results, achieving good repeatability Effect

Active Publication Date: 2019-06-28
北京羽医甘蓝信息技术有限公司
View PDF6 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In the case of missing teeth, especially for a long time, it is difficult to manually identify the tooth position
What's more, my country's medical resources are relatively scarce, and the number of senior medical experts is insufficient. If relying entirely on manual work, the operator's eyes will be tired, and the accuracy and credibility of the judgment results will be further reduced.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A panorama constant-tooth identification method and device based on deep learning
  • A panorama constant-tooth identification method and device based on deep learning
  • A panorama constant-tooth identification method and device based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0020] figure 1 It is a schematic flowchart of a method for identifying permanent teeth based on deep learning in an embodiment of the present invention, including the following steps A to D.

[0021] Step A: Input the original panorama into the alveolar bone line segmentation model based on deep learning to obtain the alveolar bone line segmentation result.

[0022] Optionally, the alveolar bone line segmentation model based on deep learning in step A is obtained in the following manner: (1) Model design stage: the network structure includes an encoder part and a decoder part, and the encoder part adopts the Xception network structure , the decoder part adopts the FPN multi-scale fusion structure, and extracts five convolutional layers with different depths from the encoder part as input; (2) Model training stage: normalize the size of the alveolar bone line segmentation training data, gray Degree normalization, data enhancement, and data balance processing. The parameters o...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a panorama constant-tooth identification method and device based on deep learning. The method comprises the following steps: inputting an original panoramic film into an alveolar bone line segmentation model based on deep learning to obtain an alveolar bone line segmentation result; generating an attached rectangular frame according to the alveolar bone segmentation result,generating an extended rectangular frame according to the attached rectangular frame, and cutting out periodontal region image blocks from the original panoramic sheet according to the extended rectangular frame; inputting the periodontal region image block into a constant tooth segmentation model based on deep learning to obtain a constant tooth segmentation result; and carrying out tooth numbermarking on each constant tooth in the constant tooth segmentation result.

Description

technical field [0001] The present invention relates to the technical field of computer image processing, in particular to a method and device for identifying permanent teeth in panoramic images based on deep learning. Background technique [0002] Oral panoramic film is the main basis for oral diagnosis, which can clearly and completely display the whole picture of maxilla, mandible, upper and lower jaw dentition, alveolar bone, maxillary sinus cavity, sinus wall, sinus floor and temporomandibular Joint conditions, and provide accurate and effective help for the diagnosis of diseases around the jaw. [0003] The segmentation of permanent teeth and the identification of tooth marks are essential for the subsequent assessment of dental health and diagnosis of diseases. figure 1 Standard tooth numbering is shown. [0004] In the case of missing teeth, especially when the tooth has been missing for a long time, it is difficult to manually identify the tooth position. What's ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/11G06T7/00
Inventor 徐子能白海龙丁鹏汪子晨
Owner 北京羽医甘蓝信息技术有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap