Higher level brain glioma survival period prediction method based on sparse representation framework and higher level brain glioma survival period prediction system based on sparse representation framework

A sparse representation, glioma technology, applied in the field of medical image processing, can solve problems such as limiting the comprehensive utilization of deep pathological features

Active Publication Date: 2018-07-24
FUDAN UNIV
View PDF6 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most of these methods use manual features or engineering features guided by previous clinical experience for research and analysis, which limits the comprehensive utilization of deep pathological features contained in biomedical 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
  • Higher level brain glioma survival period prediction method based on sparse representation framework and higher level brain glioma survival period prediction system based on sparse representation framework
  • Higher level brain glioma survival period prediction method based on sparse representation framework and higher level brain glioma survival period prediction system based on sparse representation framework
  • Higher level brain glioma survival period prediction method based on sparse representation framework and higher level brain glioma survival period prediction system based on sparse representation framework

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0034] The following are the specific implementation steps of the whole method:

[0035] 1. Firstly, perform operations such as brain removal and gray-scale normalization on the images in the data set, and select 30 images from the T1-enhanced and T2-weighted MRI image collections for manual labeling of tumor areas, and then send the labeling results and corresponding images to Two kinds of convolutional neural networks constructed were used to train the network parameters, and finally the two trained convolutional neural networks were used to segment the tumor area of ​​the corresponding modality image.

[0036] 2. Extract the local SIFT features of the two modalities of each patient, and the key points of the SIFT features are evenly distributed in each layer of the tumor image at intervals of 8 pixels. For each key point, first divide the 16*16 pixel neighborhood into 16 4*4 pixel sub-neighborhoods, then calculate the gradient size and direction of each sub-neighborhood, an...

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 belongs to the technical field of medical image processing, and particularly provides a higher level brain glioma survival period prediction method based on a sparse representation framework and a higher level brain glioma survival period prediction system based on the sparse representation framework. Firstly the tumor area in an MRI image is segmented by using an image segmentationmethod based on a convolutional neural network; then the global features of the tumor area are extracted by using the method based on local SIFT feature sparse representation; feature selection is performed by using the model of combining sparse representation and structure preserving for reducing the feature redundancy and enhancing the survival period prediction accuracy, and a few high-stability and high-resolution features are selected to predict the survival period; and finally whether the survival period of the patient is greater than 22 months is predicted by using a multi-feature collaboration sparse representation classification method with combination of the multimodal features. The method has high survival period prediction accuracy, and the whole process is automatically completed by the computer without manual intervention so that the method can be used for postoperative survival period prediction of the clinical higher level brain glioma.

Description

technical field [0001] The invention belongs to the technical field of medical image processing, and in particular relates to a multimodal MRI image processing method and system, in particular to a method and system for predicting the survival period of higher-level brain gliomas. Background technique [0002] The World Health Organization (WHO) defines higher-grade gliomas as III and IV grades not only have poor prognosis clinically, but also have a large difference in overall survival between different patients. Accurate postoperative survival prediction can provide optimal guidance for tumor treatment plan, which has important clinical research value. WHO pathological grade, image features and some basic clinical information including age and sex are widely used in survival prediction research. However, most of these methods use manual features or engineering features guided by previous clinical experience for research and analysis, which limits the comprehensive utiliza...

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): G06K9/62G06K9/46G06T7/00G06T7/11
CPCG06T7/0012G06T7/11G06T2207/10088G06T2207/30096G06V10/513G06V10/462G06F18/24G06F18/214
Inventor 余锦华汪源源吴国庆
Owner FUDAN UNIV
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