Multi-class medical image judgment method and system

A medical image and image acquisition technology, which is applied in the field of image processing, can solve the problems of reduced work efficiency, difficult identification, and taking too much time, so as to reduce the processing burden and improve the processing speed.

Active Publication Date: 2019-04-16
HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL
View PDF8 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At the same time, the number of skilled physicians will not increase significantly. In this case, identifying medical images will take too much time for doctors, resulting in reduced work efficiency, which is not conducive to medical diagnosis and increases the probability of medical accidents.
[0003] At the same time, there are many types of medical images, and different types of images have different characteristics. It is very difficult to master the recognition of all or most of the images. The amount of image recognition training is inefficient and expensive, so there is a need for a method that can effectively reduce the number of physicians judging medical images.

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
  • Multi-class medical image judgment method and system
  • Multi-class medical image judgment method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0020] This embodiment is used to illustrate the shortcomings of the prior art and the solution ideas of the present invention

[0021] It is a reasonable choice to detect the internal organs in the body through microwave imaging, because it will not cause damage to the internal organs, nor will it cause traces on the body surface; from this, various methods of obtaining the health status of the human body with the help of medical images are derived, and realistic Proving that these methods are effective has been accompanied by an increase in the amount of image data resulting from the wider use of medical images.

[0022] In the case that artificial intelligence still fails to achieve high-resolution capabilities, in fact, the final confirmation of the disease still needs to be carried out by human doctors, which brings a lot of work burden to doctors, and physical fatigue will reduce the recognition ability. This will bring great disadvantages to the diagnosis of the disease...

Embodiment 2

[0035] This embodiment is used to explain the preset rules on the basis of Embodiment 1. The purpose of the preset rules is to selectively exclude some content / data during image processing. The specific rules include:

[0036] 1) Regional area: By analyzing the tumor diameter in the MIAS data set (that is, the William Consing breast cancer data set), it can be judged that the possible area of ​​the tumor is within a certain range, and the area beyond this range is likely to be the pectoral muscle area;

[0037] 2) Shape: According to the smallest outlying rectangle of the candidate area, the aspect ratio of the rectangle is used as the shape factor, and if it exceeds a certain size, it can be removed;

[0038] 3) Average gray value: the pixel value of the tumor area is generally relatively high, and if it is lower than a certain threshold, it can be determined that the area is not a tumor;

[0039] 4) Grayscale variance: the variance of the tumor area is generally not large (t...

Embodiment 3

[0047] This embodiment provides as figure 2 A medical image processing framework is shown, including:

[0048] As the source of medical images, the medical image database extracts various features from medical images (including color moment, gray level co-occurrence matrix, directional gradient histogram and local binary mode, etc.), and fuses various features through random forest features, based on The classification processing of the SVM classifier, the category of the output image (that is, the category label), according to the category label, selects an appropriate processing algorithm from the disease image processing algorithm database to process the medical image and obtain various features (including color moments, gray level co-occurrence matrix) , directional gradient histogram and local binary pattern, etc., at this time, mark it as a disease feature), and then classifier (any combination) according to KNN (K nearest neighbor), SVM (support vector machine), BPNN (...

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 multi-class medical image judgment method and system. The method comprises the steps of processing medical images to obtain image features, analyzing different kinds of medical images, and obtaining an optimal feature subset based on a random deep forest classifier; processing the optimal feature subset through a support vector machine to output a classification result and endow the classification result with a type label corresponding to the medical image; and selecting a corresponding disease image processing algorithm based on the type label to process the medicalimage again to obtain disease characteristics, and processing the disease characteristics based on a classification network to output a diagnosis result. The system is for executing corresponding method. The method comprises the following steps: processing a medical image to obtain image features, and obtaining an optimal feature subset based on a classifier. The classification result is output through the support vector machine, the disease image processing algorithm is selected according to the classification result to process the medical image to obtain the disease characteristics, the disease characteristics are processed based on the classification network to output the diagnosis result, the medical image processing speed can be increased, and the processing burden of doctors is reduced.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a multi-type medical image judgment method and system. Background technique [0002] In the clinic, most of the medical data are medical images, and with the update of image acquisition technology and the reduction of equipment costs, the importance of medical images will increase in the medical field in the future, and the number generated in the diagnosis process will also increase. At the same time, the number of skilled doctors will not increase significantly. In this case, identifying medical images will take too much time for doctors, resulting in reduced work efficiency, which is not conducive to medical diagnosis and increases the probability of medical accidents. [0003] At the same time, there are many types of medical images, and different types of images have different characteristics. It is very difficult to master the recognition of all or most of the imag...

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): G06K9/62G06K9/32G06K9/34G06K9/40G16H50/20G06N3/04G06N3/08
CPCG06N3/08G16H50/20G06V10/30G06V10/25G06V10/267G06N3/044G06N3/045G06F18/2411G06F18/214Y02A90/10
Inventor 曾国坤王蓝天曾品榛
Owner HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL
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