Artificial intelligence capsule endoscopy examination method and system based on deep reinforcement learning

A technology of intensive learning and capsule endoscopy, which is applied in the direction of endoscopy, gastroscope, esophagus, etc., can solve the problems of difficult complete shooting, fast speed and large volume of capsule endoscopy, and achieve good prognosis, easy acceptance and saving The effect of time cost

Active Publication Date: 2018-11-13
WUHAN ENDOANGEL MEDICAL TECH CO LTD
View PDF2 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the stomach is a hollow organ with a large volume, and the capsule endoscope passes through quickly, so it is difficult to take a complete image of the above-mentioned parts, and there are many observation blind spots; the magnetically controlled capsule endoscope is a kind of remote control that the doctor pushes

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
  • Artificial intelligence capsule endoscopy examination method and system based on deep reinforcement learning
  • Artificial intelligence capsule endoscopy examination method and system based on deep reinforcement learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0044] The artificial intelligence capsule endoscopy method based on deep reinforcement learning of the embodiment of the present invention, the method includes:

[0045] Step S10: The capsule endoscope collects images through the camera, processes the images, extracts image feature values, and takes this as the pre-exercise state;

[0046] The image taken by the capsule endoscope is used as the current state of the agent, which is input into the deep reinforcement learning module in order to obtain a decision-making action. In this embodiment, in order to reduce computer processing time, the collected images can be preprocessed, including performing operations such as grayscale and downsampling, and then input the preprocessed images into the convolutional neural network for feature value extraction.

[0047] Step S20: Input the image feature value extracted in step S10 into the pre-trained deep reinforcement learning model to obtain the maximum value action that the capsule ...

Embodiment 2

[0068] On the other hand, the artificial intelligence capsule endoscopy system based on deep reinforcement learning of the embodiment of the present invention includes:

[0069] State acquisition module: used to obtain the current state of the capsule endoscope according to the images captured by the capsule endoscope camera;

[0070] Deep reinforcement learning module: used to input the feature value into the deep reinforcement learning model for processing to obtain the action corresponding to the maximum value; used to input the training data into the module and train the depth reinforcement learning model so that the trained depth reinforcement Learning models enable decision analysis;

[0071] Control instruction generation and execution module: use the maximum value action output by the deep reinforcement learning module to generate corresponding control instructions according to the state of the capsule endoscope, adjust the voltage of the gradient coil and shim coil ou...

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 an artificial intelligence capsule endoscopy examination method and an artificial intelligence capsule endoscopy examination system based on deep reinforcement learning. The method comprises the following steps: acquiring an image of the gastral cavity environment by adopting a capsule endoscopy, and extracting characteristic values of the image; inputting the characteristic values of the image into a pre-trained deep reinforcement learning model, thus obtaining maximum value movements capable of being executed; by utilizing the maximum value movements output by the deep reinforcement learning model, generating a corresponding control instruction according to the state of the capsule endoscopy, and controlling the capsule endoscopy to carry out autokinetic movementin the complicated gastral cavity environment; after the movements of the autokinetic movement are completed, acquiring a return value according to the actual completion status; and judging that whether the capsule endoscopy arrives at a final position or not. The method and the system aim at realizing the purposes that through training, the capsule endoscopy can make the correct decision in the complicated and high-dynamic gastral cavity environment, the capsule endoscopy can be controlled to carry out autokinetic movement in the complicated gastral cavity environment, and the examination forthe whole stomach is intelligently and efficiently realized without omission.

Description

technical field [0001] The invention relates to the field of medical devices, in particular to an artificial intelligence capsule endoscopy method and system based on deep reinforcement learning. Background technique [0002] my country is a big country with stomach problems, especially gastric cancer, the morbidity and mortality have been high. In the prior art, endoscopy is of great significance in the diagnosis of digestive system diseases. [0003] As an invasive examination, the traditional electronic gastroscope will irritate the patient's throat, bring physiological discomfort to the patient, reduce the patient's compliance, and delay the early diagnosis of the disease; in addition, the doctor's operating level and experience are relatively high. High, it takes a long time to train an excellent digestive endoscopist doctor; at the beginning of the new century, capsule endoscopy came to the stage. Patients only need to swallow a small capsule, and they can go through...

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): A61B1/045A61B1/273G06K9/62
CPCA61B1/00006A61B1/00009A61B1/041A61B1/045A61B1/2736G06F18/214
Inventor 于红刚吴练练宫德馨
Owner WUHAN ENDOANGEL MEDICAL TECH CO LTD
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