Unlock instant, AI-driven research and patent intelligence for your innovation.

Pulmonary nodule detection model training method and device and pulmonary nodule detection method and device

A technology for detecting models and training methods, applied in image data processing, instruments, computing, etc., can solve the problems of high labeling cost, difficulty in labeling pulmonary nodules, and high requirements for labelers

Pending Publication Date: 2020-12-25
GUANGZHOU SHIYUAN ELECTRONICS CO LTD
View PDF5 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, it is extremely difficult to label pulmonary nodules in CT images of the lungs, and the requirements for the labeler are high, that is, the labeling cost is very high

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
  • Pulmonary nodule detection model training method and device and pulmonary nodule detection method and device
  • Pulmonary nodule detection model training method and device and pulmonary nodule detection method and device
  • Pulmonary nodule detection model training method and device and pulmonary nodule detection method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0044] figure 1 It is a flow chart of a pulmonary nodule detection model training method provided by Embodiment 1 of the present invention. This embodiment can be applied to the deep learning process, where it is difficult to label CT images of the lungs, and the labeling takes a long time and requires high requirements for the labeler. In this case, the method can be executed by the pulmonary nodule detection model training device provided by the embodiment of the present invention. The device can be implemented by software and / or hardware, and is usually configured in a computer device, such as figure 1 As shown, the method specifically includes the following steps:

[0045] S101. Train a pulmonary nodule detection model using labeled data as samples.

[0046] Specifically, a data set is obtained first, and the data set includes labeled labeled data and unlabeled unlabeled data. Wherein, the labeled data includes a plurality of first lung CT image samples with labels, and ...

Embodiment 2

[0069] Figure 2A It is a flow chart of a pulmonary nodule detection model training method provided by Embodiment 2 of the present invention. This embodiment is refined on the basis of the above-mentioned Embodiment 1, and describes in detail the network structure and processing of the pulmonary nodule detection model process, as well as the structure of the graph convolutional neural network and its processing, such as Figure 2A As shown, the method includes:

[0070] S201. Train a pulmonary nodule detection model using labeled data as samples.

[0071] Specifically, a data set is obtained first, and the data set includes labeled labeled data and unlabeled unlabeled data. Wherein, the labeled data includes a plurality of first lung CT image samples with labels, and the labels are used to indicate whether there are pulmonary nodules in the first lung CT image samples. The unlabeled data includes multiple unlabeled second lung CT image samples.

[0072] In a specific embod...

Embodiment 3

[0175] image 3 It is a flow chart of a pulmonary nodule detection method provided in Embodiment 3 of the present invention. The method uses the pulmonary nodule detection model trained by the pulmonary nodule detection model training method provided in any of the above embodiments for prediction. The method It can be performed by the pulmonary nodule detection device provided by the embodiment of the present invention, which can be implemented by software and / or hardware, and is usually configured in a computer device, such as image 3 As shown, the method specifically includes the following steps:

[0176] S301. Acquire a CT image of the lung to be detected.

[0177] Specifically, in the embodiment of the present invention, the lung CT image may be a two-dimensional CT image or a three-dimensional CT image. Exemplarily, in a specific embodiment, the lung CT image is a three-dimensional CT image.

[0178] S302. Input the lung CT image into the pulmonary nodule detection mod...

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 pulmonary nodule detection model training method and device and a pulmonary nodule detection method and device. The pulmonary nodule detection model is trained based on the semi-supervised learning method, the number of annotation data needed for training the pulmonary nodule detection model is reduced, and then the annotation cost is reduced. Furthermore, by calculatinga connection matrix between the nodule features, constructing a graph convolutional neural network based on the connection matrix, and fully mining common features between the labeled data and the unlabeled data by using the graph convolutional neural network, the detection precision of the pulmonary nodule detection model can be improved.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of machine learning, and in particular to a pulmonary nodule detection model training method, detection method and device. Background technique [0002] Affected by factors such as environment, smoking and genetics, lung cancer is the malignant tumor with the highest mortality and morbidity in my country. According to medical data, the 5-year survival rate of early lung cancer is significantly higher than that of advanced lung cancer. Early detection, early diagnosis and treatment are important ways to improve lung cancer. [0003] Judging whether there are pulmonary nodules in the lungs is a powerful indicator for judging cancer, so early screening of pulmonary nodules becomes particularly important. Among them, low-dose chest CT images have the characteristics of thin layers, clear field of vision, and few interference factors. Therefore, the detection of pulmonary nodules based on l...

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): G06T7/00
CPCG06T7/0012G06T2207/10081G06T2207/20081G06T2207/20084G06T2207/30064Y02A90/10
Inventor 张玉兵王静雯
Owner GUANGZHOU SHIYUAN ELECTRONICS CO LTD