A method for predicting the aggressiveness of thyroid nodules based on deep learning segmentation network

A thyroid nodule and deep learning technology, applied in the field of image processing, can solve the problems of insufficient accuracy and low detection rate, and achieve the effects of improving accuracy, increasing detection speed, and overcoming poor image quality

Active Publication Date: 2022-03-04
INNER MONGOLIA UNIVERSITY
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In view of the problems existing in the existing technical solutions, the purpose of the present invention is to provide a method for predicting the aggressiveness of thyroid nodules based on a deep learning segmentation network, which can overcome the defects of insufficient accuracy and low detection rate of traditional detection methods

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 method for predicting the aggressiveness of thyroid nodules based on deep learning segmentation network
  • A method for predicting the aggressiveness of thyroid nodules based on deep learning segmentation network
  • A method for predicting the aggressiveness of thyroid nodules based on deep learning segmentation network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0073] This embodiment provides a method for predicting the aggressiveness of thyroid nodules based on a deep learning segmentation network, such as figure 1 As shown, the method includes the following steps:

[0074] S1: The adaptive wavelet algorithm is used to preprocess the clinically obtained thyroid ultrasound images, remove image noise and retain the edge information of the images in the high-frequency domain, and obtain the original data set.

[0075] Aiming at the problems of poor image quality, severe speckle noise, fuzzy edges of nodules, discontinuous boundaries, low contrast, and concentrated edge information and severe noise in the high-frequency domain. In this embodiment, an adaptive wavelet algorithm is used to preprocess the image. Wavelet filtering is based on the wavelet change, transforming the signal in the spatial domain to the wavelet domain with time-frequency characteristics, and then using the wavelet coefficients mapped by the threshold to reduce n...

Embodiment 2

[0125] like Figure 10 As shown, this embodiment provides a thyroid nodule invasiveness prediction system based on a deep learning segmentation network. To predict the conclusion of nodule invasiveness, the system includes:

[0126] A preprocessing module, which is used for preprocessing the clinically obtained thyroid ultrasound image according to the method in embodiment 1, removing image noise and retaining edge information of the image in the high frequency domain;

[0127]Generate an adversarial network module, which includes a generator submodule and a discriminator submodule; the generated adversarial network module uses the method in Embodiment 1 to carry out semantic segmentation of nodule instances on thyroid ultrasound images, obtains nodule masks, and then extracts Output the nodule area information, edge information and aspect ratio information; and binarize the mask output by the generator module, and then multiply it with the original image to obtain the image ...

Embodiment 3

[0130] This embodiment provides a terminal for predicting the aggressiveness of thyroid nodules based on a deep learning segmentation network. The terminal includes a memory, a processor, and a computer program stored in the memory and operable on the processor. The processor The method for predicting the aggressiveness of thyroid nodules based on the deep learning segmentation network as in Example 1 was implemented.

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 belongs to the technical field of image processing, and specifically relates to a method for predicting the aggressiveness of thyroid nodules based on a deep learning segmentation network. The method includes the following steps: S1: Preprocessing the thyroid ultrasound images obtained clinically; S2: Constructing the main structure framework based on the deep learning segmentation network; S3: Improving the generation confrontation network model in the main structure framework; S4: Thyroid nodules are accurately semantically segmented, and the information of nodule area, aspect ratio, and contour regularity is calculated; S5: Obtain a new image dataset that only contains nodules after cropping; S6: Non-destructive classification of the classification network model Improve the linear expression ability, S7: Use the improved classification network model to classify the prediction results, and train and update the classification network model. The method provided by the invention can realize end-to-end automatic auxiliary diagnosis, and overcome the defects of insufficient accuracy and low detection rate of traditional detection methods.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a method for predicting the aggressiveness of thyroid nodules based on a deep learning segmentation network. Background technique [0002] The thyroid is the largest endocrine gland in the human body. Ultrasound examination can make qualitative and quantitative estimates of its size, volume and blood flow, and can make qualitative or semi-qualitative diagnosis of benign and malignant tumors. Therefore, ultrasonic detection methods have also become imaging examinations. The preferred method for thyroid disease. In the past, the results of thyroid ultrasound examination were mainly judged by doctors based on experience, and conclusions were predicted. After the introduction of image recognition technology, various detection systems based on classification can replace manual processing and prediction of ultrasound image data, thereby greatly improving the detection efficie...

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 Patents(China)
IPC IPC(8): G06T7/00G06T7/12G06T7/62G06T5/00G06V10/44G06V10/764G06K9/62
CPCG06T7/0012G06T7/12G06T5/002G06T7/62G06T2207/10132G06T2207/20081G06T2207/20084G06T2207/20192G06T2207/30096G06T2207/20132G06V10/44G06F18/24
Inventor 郑志强王雨禾翁智
Owner INNER MONGOLIA UNIVERSITY
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