Femur neck fracture detection method and system based on weak supervision segmentation

A femoral neck fracture and weakly supervised technology, which is applied in the fields of radiological diagnostic instruments, medical science, radiological diagnostic image/data processing, etc., can solve the problems of femoral neck area detection, low efficiency, and poor detection accuracy

Active Publication Date: 2020-09-29
浙江飞图影像科技有限公司
View PDF10 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The present invention provides fracture detection for the prior art, but does not specifically detect the femoral neck region; for femoral neck fracture detection, it detects on the basis of multi-model fusion, and its detection accuracy is poor and the efficiency is low Shortcomings, providing a femoral neck fracture detection method and system based on weakly supervised segmentation

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
  • Femur neck fracture detection method and system based on weak supervision segmentation
  • Femur neck fracture detection method and system based on weak supervision segmentation
  • Femur neck fracture detection method and system based on weak supervision segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0075] The method for femoral neck fracture detection based on weakly supervised segmentation includes the following detection steps,

[0076] Obtain the detection image, input the obtained femoral neck X-ray film into the constructed and trained detection neural network to detect the femoral part, and obtain the probability segmentation map P probability and femoral neck area detection image P detec ;

[0077] Obtain the fracture type of the femoral neck region, and detect the femoral region image P detec Input to the constructed and trained classification network to obtain the type of femoral fracture;

[0078] Obtain the segmented fracture site, and detect the femoral region image P detec Input the constructed and trained weakly supervised segmentation network to obtain the segmented fracture image P seg ;

[0079] Image fusion is used to obtain the image of the femoral neck fracture area, and the segmented fracture site image P seg with femoral region detection image...

Embodiment 2

[0124]On the basis of Embodiment 1, the system of femoral neck fracture detection based on weak supervision segmentation in this embodiment includes a detection module, a classification module, a weak supervision segmentation module and an image fusion module, and is characterized in that it also includes a femoral fracture detection system based on weak supervision segmentation. Methods of neck fracture detection,

[0125] The detection module inputs the obtained X-rays of the femoral neck into the constructed and trained detection neural network to detect the femoral part, and obtains the probability segmentation map P probability and femoral neck area detection image P detec ;;

[0126] Classification module, the femoral region detection image P detec Input to the constructed and trained classification network to obtain the type of femoral fracture;

[0127] The weakly supervised segmentation module obtains the segmented fracture site and detects the femoral region image...

Embodiment 3

[0130] On the basis of embodiment 1, the steps of constructing neural network and segmentation probability map in this embodiment include,

[0131] Step 1.1, input a set of femoral neck X-ray images P orig , set the initial convolution output channel 64;

[0132] Step 1.2, for the input image P orig Do 2 convolutions of 3×3, and the number of output channels for each convolution is n 2 , batch normalization and ReLU operations are performed after each convolution, and a 2×2 maximum pooling operation is performed after the two convolutions are completed to extract the feature EIG_S2;

[0133] Step 1.3, use the feature EIG_S2 as the input feature, perform two 3×3 convolutions, and the number of output channels for each convolution is n 3 , n 3 =n 2 *2; when n 3 =512, the output feature is EIG_S3, and step 1.4 is performed, otherwise, the output feature EIG_S3 is used as the input feature of step 1.2, and step 1.2 is repeated;

[0134] Step 1.4, perform two 3×3 convolution...

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 relates to the field of medical image processing, and discloses a femur neck fracture detection method and system based on weak supervision segmentation; the method comprises the steps:obtaining a detection image, inputting an obtained femur neck X-ray film into a constructed and trained detection neural network, and carrying out the detection of a femur part; obtaining a probability segmentation map P<probability and a femur neck region detection image P<detec>; acquiring the fracture type of the femur neck region, inputting the femur region detection image P<detec> into a constructed and trained classification network, and acquiring the type of the femur part fracture; acquiring a segmented fracture part, and inputting the femur region detection image P<detec> into a constructed and trained weak supervision segmentation network to obtain a segmented fracture part image P<seg>; and performing image fusion to obtain a femur neck region fracture image, and performing fusing on the segmented fracture part image P<seg> and the femur region detection image P<detec>to generate and output a femoral neck fracture region image. The femoral neck fracture is segmented througha weak supervision segmentation method, the detection precision is high, and the efficiency is high.

Description

technical field [0001] The invention relates to the field of medical image processing, in particular to a femoral neck fracture detection method and system based on weak supervision segmentation. Background technique [0002] Femoral neck fracture is a common clinical trauma, accounting for about 3.58% of all fractures. Femoral neck fractures often occur in the elderly, and its incidence is increasing with the increase of life expectancy, especially in developed countries and regions where the elderly population is increasing rapidly. According to statistics, the number of hip fractures in the world was 10,000 in 1990, and this number is expected to reach 4 million by 2025, and 6.3 million by 2050. The mortality rate and morbidity rate of femoral neck fracture injury are high, and fracture nonunion and avascular necrosis of femoral head are two main problems existing in its clinical treatment. [0003] Early diagnosis and treatment can not only maintain and reduce the morb...

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): A61B6/00
CPCA61B6/505A61B6/52
Inventor 胡利荣符勇张跃华
Owner 浙江飞图影像科技有限公司
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