Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for processing head structure image data based on deep learning algorithm

A technology of deep learning and head processing, which is applied in the field of medical image processing, can solve problems such as applied research and prediction, and achieve the effects of reducing work pressure, improving work efficiency and reducing work pressure

Inactive Publication Date: 2016-12-21
王双坤 +2
View PDF0 Cites 53 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the application research and prediction of deep learning methods in the diagnosis and treatment of clinical medical diseases are still rare, especially in the human brain. This invention applies the deep learning method to the discrimination of the brain structure of smokers for the first time, and explores a clinically proven method. and a new way to identify structural changes in the brain of smokers

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
  • Method for processing head structure image data based on deep learning algorithm
  • Method for processing head structure image data based on deep learning algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0037]The head magnetic resonance structural image data includes head structural magnetic resonance images of smokers and non-smokers, and the parameters of the head structural magnetic resonance images are: 3D T1-weighted MPRAGE T1 weighted sequence, sagittal scan, 176 layers , TR=2530ms, TE=3.39ms, slice thickness 1mm, TI=1100ms, FOV=256mm×256mm, resolution=256×256, repeated twice.

[0038] Such as figure 1 As shown, there are 176 MRI images of head structures in each case, and 1 / 4 of the smoking group and non-smoking group were randomly selected as the test set, including 15 smokers and 16 non-smokers, and the rest were used as the training set.

[0039] Efficiently handle image tasks through deep convolutional network models and recurrent neural networks (RNN). As a long-short-term memory recurrent neural network, LSTM can map variable-length inputs into variable-length outputs.

[0040] figure 2 shows the flowchart of the ConvLSTM model, as figure 2 As shown, firstl...

Embodiment 2

[0043] In Example 2, there are also 176 images of the MRI image of head structure in each case. The drug-taking group and the non-drug-taking group each randomly select 1 / 4 as the test set, including 15 cases of drug users, 16 cases of non-drug users, and the rest as the training set. .

[0044] All the other methods are the same as in Example 1.

Embodiment 3

[0046] In Example 3, there are also 176 images of head structure magnetic resonance images in each case, and each of the epileptic patient group and the non-epilepsy patient group randomly selects 1 / 4 as the test set, including 15 epileptic patients and 16 non-epilepsy patients. example, and the rest as the training set.

[0047] All the other methods are the same as in Example 1.

[0048] The method for processing head structure image data based on deep learning algorithm of the present invention uses LSTM model to analyze head structure image data, so as to better achieve classification effect.

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 method for processing head structure image data based on a deep learning algorithm. The method comprises the following steps of acquiring head magnetic resonance structure image data and storing the head magnetic resonance structure image data in a computer; first extracting corresponding features of multiple frames of pictures through convolutional neural network models respectively, then taking a feature of a sequence picture as a sequence input of an LSTM (Long Short-Term Memory) model, and taking combination of the feature of the sequence picture and a tag corresponding to a whole sequence as a training sample of the LSTM model; sequentially inputting 100 layers of pictures in the middle of an MRI (Magnetic Resonance Imaging) image by adopting 3 layers of LSTM models, wherein each layer of picture corresponds to a 1024-dimensional feature, and the end of the model is a full connection layer of two classifiers; importing to-be-analyzed medical data into a deep learning model to perform matched medical analysis; and when the model training is finished, arranging test samples in the same format as an MRI sequence layer, and then performing classification by using the trained deep learning model to judge whether the MRI image belongs to a smoker, a drug addict or a disease patient.

Description

technical field [0001] The invention belongs to the technical field of medical image processing, and in particular relates to a method for processing head structure image data based on a deep learning algorithm. The method is based on a deep convolutional neural network and a recursive neural network model. Background technique [0002] Nicotine dependence has been defined as a disease by the World Health Organization and included in the International Classification of Diseases (ICD-10, F17.2) (WHO1992), which belongs to psychoactive substance dependence, and its occurrence and development are related to changes in the structure and function of the brain. close relationship. Smoking addiction is not only a physical dependence, but also a psychological dependence, which not only has a significant impact on human brain cognition and attention functions, but also leads to changes in brain structure and function, and is also a risk factor for cerebrovascular diseases and dementi...

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): G06F19/00G06K9/46G06K9/62
CPCG16H50/20G06V10/40G06F18/22
Inventor 王双坤陈宽张荣国
Owner 王双坤
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com