Method and system for automatic monitoring of retention of digestive tract capsule endoscope
A capsule endoscope and automatic monitoring technology, applied in the field of image processing, can solve problems such as detection of digestive tract interference content and capsule endoscope retention monitoring without taking into account, and achieve the effect of convenient monitoring and improved accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0035] like figure 1 As shown, an automatic monitoring method for gastrointestinal capsule endoscopy retention is provided, including the following contents:
[0036] Build an interference image detection network model, obtain interference image data, perform network training on the interference image detection network model, and remove the interference images;
[0037] Build an organ detection network model, obtain the organ image data after removing the interference images, train the organ detection network model and output the organ category judgment information;
[0038] The position and state of the capsule endoscope are determined by combining the organ type determination information and the time information of swallowing the capsule endoscope.
[0039] The principle of the method is as follows: firstly, a network model for interference image detection is constructed, and the interference image data is obtained to perform network training on the interference image detec...
Embodiment 2
[0046] like figure 2 As shown, an automatic monitoring system for gastrointestinal capsule endoscopy retention is provided, and the system includes:
[0047] Image data acquisition module: acquires the interference image data and the organ image data after removing the interference image;
[0048]Model training module: including an interference image detection network model training module and an organ detection network model training module; the interference image detection network model training module is used to construct an interference image detection network model, obtain interference image data, and perform network operation on the interference image detection network model. Train and remove interference images, and output the organ images after interference removal through the interference content detection module.
[0049] The organ detection network model training module is used to construct an organ detection network model, obtain organ image data after removing i...
Embodiment 3
[0055] On the basis of Embodiments 1 and 2, a preferred embodiment is provided, such as Figure 3-5 As shown, when acquiring data, the interference image data is obtained through the image data acquisition module. The types of the interference image data include air bubbles in the digestive tract, intestinal fluid and food residues, etc. The interference image data includes the interference image data training set and the interference image data validation set.
[0056] Further, the interference image data training set includes clear training pictures of digestive tract mucosa, clear training picture labels of digestive tract mucosa, interference training pictures, and interference training picture labels, and the interference image data verification set includes clear verification pictures of digestive tract mucosa. , Clear verification picture mark of digestive tract mucosa, interference verification picture, interference verification picture mark.
[0057] Further, a networ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


