Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Thyroid nodule ultrasonic image classification method based on capsule network

A technology for thyroid nodules and ultrasound images, which is applied in image analysis, image enhancement, graphics and image conversion, etc. It can solve the problems of no translation of the model and loss of important information, etc., and achieve high accuracy and stable fluctuation of the accuracy change curve Effect

Pending Publication Date: 2019-07-26
TIANJIN UNIV
View PDF5 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0018] The invention provides a capsule network-based ultrasound image classification method for thyroid nodules. The invention effectively overcomes the loss of important information in the pooling process of the traditional convolutional neural network classification technology, and the trained model does not have translation and rotation invariance, etc. Problem, improve the accuracy of the capsule network in the thyroid ultrasound image classification task, see the description below for details:

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
  • Thyroid nodule ultrasonic image classification method based on capsule network
  • Thyroid nodule ultrasonic image classification method based on capsule network
  • Thyroid nodule ultrasonic image classification method based on capsule network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0048] The embodiment of the present invention provides a method for classifying ultrasound images of thyroid nodules based on a capsule network, see figure 1 , the method includes the following steps:

[0049] 101: Preprocessing the ultrasound images of the thyroid gland to identify artificial markers in the ultrasound images;

[0050] 102: Obtain a rectangular frame for defining the position of the thyroid nodule through boundary adjustment;

[0051] 103: Construct a capsule network and apply it to the classification of ultrasound images of thyroid nodules, and adjust the structure of the capsule network;

[0052] 104: Add the Dropout method to the adjusted capsule network to stabilize the training process and improve the classification effect.

[0053] In one embodiment, the ultrasound image of the thyroid gland is preprocessed through step 101, and the specific steps are as follows:

[0054] In ultrasound images, a pair of "+" symbols (plus sign) are usually used to rep...

Embodiment 2

[0066] The scheme in embodiment 1 is further introduced below in conjunction with specific calculation formulas and examples, see the following description for details:

[0067] 201: Using image differentiation to identify artificial marks;

[0068] Among them, the Laplacian operator is a second-order differential operator. Applying a Laplacian operator L to the image R can obtain the second-order differential of the image R. In order to facilitate further processing, the second-order differential The obtained image is binarized and recorded as image G. The process is shown in formula (1), where η is the threshold used for image binarization, and the transformed image G is a binary image that only contains two gray values ​​of 0 and 255. Among them, C=R×L.

[0069]

[0070] 202: The structure of the neural network used is shown in Table 1. After each convolutional layer, a ReLU is added as an activation function;

[0071] Table 1 Neural network structure for identifying ...

Embodiment 3

[0099] Combine below Figure 3-Figure 5 , and specific calculation formulas carry out feasibility verification to the scheme in embodiment 1 and 2, see the following description for details:

[0100] First of all, compared with other traditional methods, the method proposed in the embodiment of the present invention can accurately and quickly determine whether there is a manually marked symbol at a certain position in the ultrasound image, and give the category of the symbol and the probability of belonging to this category. The performance of our method compared with other methods is shown in Table 2. Combining the method of image preprocessing and border adjustment, it is possible to use a rectangular frame in the ultrasound image to locate the position of the thyroid nodule, such as image 3 shown. In the embodiment of the present invention, accuracy rate, precision rate, recall rate and F1 score are used as evaluation indicators, and the calculation formulas thereof are ...

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 discloses a thyroid nodule ultrasonic image classification method based on a capsule network, and the method comprises the following steps: carrying out the preprocessing of a thyroid ultrasonic image, and recognizing an artificial mark in the thyroid nodule ultrasonic image through the combination of a neural network and image differentiation; obtaining a rectangular frame for defining a thyroid nodule position through boundary adjustment; constructing a capsule network, applying the capsule network to classification of thyroid nodule ultrasonic images, and adjusting the structure of the capsule network; and adding Dropout into adjusted capsule network, so that training process is stabilized, and improving classification effect. According to the method, the problems that important information is lost in the pooling process and a trained model does not have translation and rotation invariance in the traditional convolutional neural network classification technology are effectively solved, and the accuracy of a capsule network in a thyroid ultrasound image classification task is improved.

Description

technical field [0001] The invention relates to the fields of deep learning, computer-aided medical treatment and medical image processing, in particular to a method for classifying ultrasound images of thyroid nodules based on a capsule network. Background technique [0002] At present, many studies apply machine learning classification methods to medical image classification and other aspects to provide effective assistance for doctors' diagnosis. In related technologies, there are mainly two methods for classifying ultrasound images of thyroid nodules: one is a traditional machine learning method. Si Luo et al. [1] A method for classifying thyroid nodules based on the characteristics of ultrasound images of thyroid nodules is proposed, by using the power spectrum of the strain rate waveform extracted from the ultrasound imaging image sequence to classify thyroid nodules, and an AUC of 0.88 is obtained in the experiment Judging score. Jieming Ma et al. [2] A SVM-based ...

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
IPC IPC(8): G06K9/62G06N3/04G06T3/00G06T5/00G06T7/11G06T7/70
CPCG06T7/11G06T7/70G06T2207/10132G06N3/045G06F18/214G06T3/02G06T5/70
Inventor 赵满坤张瑞璇魏玺刘志强于健徐天一高明张晟王臣汉喻梅于瑞国刘凯
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com