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

Duodenum self-training classification method and system based on model migration

A duodenum and classification method technology, which is applied in the field of intelligent medical image processing, can solve problems such as low detection efficiency and inaccurate classification models, and achieve the effect of improving accuracy and eliminating the influence of human factors

Active Publication Date: 2021-07-02
紫东信息科技(苏州)有限公司
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Therefore, a self-training classification method and system for duodenum based on model migration provided by the present invention overcomes the shortcomings of inaccurate classification models and low detection efficiency in the prior art due to the serious shortage of data sets.

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
  • Duodenum self-training classification method and system based on model migration
  • Duodenum self-training classification method and system based on model migration
  • Duodenum self-training classification method and system based on model migration

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] A self-training classification method for duodenum based on model migration provided by an embodiment of the present invention, such as figure 1 shown, including the following steps:

[0035] Step S1: Preprocessing the image set containing a large number of images of preset parts, extracting the feature vectors of the preset parts, inputting the feature vectors into the preset classification model for training, and obtaining the classification model of the preset parts.

[0036] In the embodiment of the present invention, the preset parts include the stomach and other organs. This is just an example and not limited thereto. In practical applications, the corresponding parts are selected according to actual needs. As an example; image preprocessing includes: rotating, scaling, translating, horizontally flipping, cropping, and normalizing the images of the preset parts and duodenum, wherein the scaling is to scale the size of all different input images to 256 The size of...

Embodiment 2

[0049] Embodiments of the present invention provide a self-training classification system for duodenum based on model migration, such as figure 2 shown, including:

[0050] The preset part classification model acquisition module 1 is used to preprocess the image set containing a large number of preset part images, extract the feature vector of the preset part, input the feature vector into the preset classification model for training, and obtain the preset part Classification model; this module executes the method described in step S1 in Embodiment 1, which will not be repeated here.

[0051] The model migration module 2 is used to transfer the trained classification model of the preset part to the duodenum by using the model migration, and obtain the transferred model. Based on a small amount of preprocessed images of normal and ulcerated duodenum, the transferred The model is trained to obtain the self-training classification model of the duodenum; this module executes the...

Embodiment 3

[0055] An embodiment of the present invention provides a terminal, such as image 3 As shown, it includes: at least one processor 401 , such as a CPU (Central Processing Unit, central processing unit), at least one communication interface 403 , memory 404 , and at least one communication bus 402 . Wherein, the communication bus 402 is used to realize connection and communication between these components. Wherein, the communication interface 403 may include a display screen (Display) and a keyboard (Keyboard), and the optional communication interface 403 may also include a standard wired interface and a wireless interface. The memory 404 may be a high-speed RAM memory (Random Access Memory, volatile random access memory), or a non-volatile memory (non-volatile memory), such as at least one disk memory. Optionally, the memory 404 may also be at least one storage device located away from the aforementioned processor 401 . Wherein, the processor 401 may execute the self-training...

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 duodenum self-training classification method and system based on model migration, and the method comprises the steps of carrying out the preprocessing of an image set containing a large number of preset part images, extracting the feature vector of a preset part, inputting the feature vector into a preset classification model, and carrying out the training, and obtaining a classification model of the preset part; migrating the trained classification model of the preset part to the duodenum by utilizing model migration to obtain a migrated model, and training the migrated model based on a small amount of preprocessed normal and ulcer images of the duodenum to obtain a duodenum self-training classification model; and pre-processing a duodenum image to be classified, inputting the pre-processed duodenum image into the duodenum self-training classification model, and calculating the probability of an image type to obtain a classification result of a normal duodenum or duodenal ulcer. According to the invention, the influence of human factors is eliminated, and the classification accuracy is effectively improved under the condition that the number of duodenum labeled samples is extremely small.

Description

technical field [0001] The invention relates to the technical field of medical image intelligent processing, in particular to a self-training classification method and system for duodenum based on model migration. Background technique [0002] Duodenal ulcer is one of the common and frequently-occurring diseases among the Chinese population, and it is a common type of peptic ulcer. It occurs mostly in winter and spring when the climate changes greatly, and the incidence rate of men is significantly higher than that of women. It is closely related to abnormal gastric acid secretion, Helicobacter pylori (Hpylori) infection, non-steroidal anti-inflammatory drugs (NSAID), irregular life and diet, work and external pressure, smoking, drinking and psychological factors. Duodenal ulcers mostly occur in the duodenal bulb, mostly in the anterior wall, followed by the posterior wall, inferior wall, and upper wall. The main diagnostic method of duodenal ulcer, gastroscopy can directly...

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
CPCG06T7/0012
Inventor 戴捷李亮
Owner 紫东信息科技(苏州)有限公司