Brain CT medical report generation method based on hierarchical recurrent neural network decoding

A technology of recurrent neural network and medical report, applied in biological neural network model, medical report, neural architecture, etc.

Pending Publication Date: 2022-03-22
BEIJING UNIV OF TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Among them, machine learning techniques have played an important role in tasks such as classification and segmentation. Howeve...

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
  • Brain CT medical report generation method based on hierarchical recurrent neural network decoding
  • Brain CT medical report generation method based on hierarchical recurrent neural network decoding
  • Brain CT medical report generation method based on hierarchical recurrent neural network decoding

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] Take the 492 examples of data provided by Peking University No. 3 Medicine as an example below to illustrate the specific implementation steps of the present invention:

[0044] Step (1) Obtain brain CT images and corresponding keywords and report data and preprocess:

[0045] Step (1.1) Collect brain CT images to build a data set, which contains a total of 492 cases of brain CT images whose sensitive information has been deleted and corresponding reports. The image data of each patient contains multiple CT sequences and a corresponding report text . The original brain CT medical image is in dicom format, which is converted into observation views of three scales commonly used by doctors as the three channel values ​​​​of the RGB three-channel color image, and the boundary noise CT value of the image -2000 is removed, and the PNG format is finally obtained. Brain CT image data I={I 1 ..., I N}, Among them, N represents the number of CT slices in each case, and W and...

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 brain CT medical report generation method based on hierarchical recurrent neural network decoding. The method comprises the steps that firstly, a brain CT image and corresponding medical report data are acquired and preprocessed; constructing a feature extractor, completing coding of the brain CT image data, obtaining coding features and fault block visual features, constructing an orientation keyword predictor, and extracting orientation keyword semantic features Fs of the brain CT image data I; constructing a hierarchical recurrent neural network language model, performing hierarchical decoding on the model by using Fs, and generating a medical report sentence by sentence; training and optimizing the model; preprocessing the brain CT to be predicted; extracting coding features and fault block visual features of the brain CT to be predicted by using the orientation keywords; extracting semantic features by using the orientation keywords; the language model generates the predicted medical report sentence by sentence using the coding features, the fault block visual features, and the semantic features.

Description

technical field [0001] The present invention uses relevant technical means in the two fields of computer vision and natural language processing, and aims at the automatic report generation task in the field of medical image analysis, and designs a medical report generation method for hierarchical cyclic neural network decoding. Background technique [0002] The task goal of automatic generation of medical reports is to input a set of brain CT images containing spatial sequence relationships, and the computer can automatically generate several sentences describing the contents of the images as corresponding medical reports, which requires the computer to have the ability to understand brain CT images , also needs to have the language organization ability to express the image content, which is a research hotspot in the current medical image analysis. [0003] With the rise of various technologies in the field of artificial intelligence, computer-assisted medical image analysis...

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
IPC IPC(8): G16H15/00G06T7/00G06K9/62G06V10/774G06V10/80G06V10/82G06N3/04
CPCG16H15/00G06T7/0012G06T2207/10081G06T2207/30016G06N3/044G06F18/214G06F18/253
Inventor 张晓丹胡启鹏刘颖王筝冀俊忠
Owner BEIJING UNIV OF TECH
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