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

A Method for Segmentation of CT Image of Vascular Plaque Based on Positional Convolutional Attention Network

A technology of vascular plaque and CT imaging, applied in the field of medical imaging, to achieve the effect of fine information, rapid screening and labeling

Active Publication Date: 2021-11-09
SHANDONG ARTIFICIAL INTELLIGENCE INST +1
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the deficiencies of the above technologies, the present invention provides a CT image segmentation method for quickly screening and labeling vascular plaques without manual intervention

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
  • A Method for Segmentation of CT Image of Vascular Plaque Based on Positional Convolutional Attention Network
  • A Method for Segmentation of CT Image of Vascular Plaque Based on Positional Convolutional Attention Network
  • A Method for Segmentation of CT Image of Vascular Plaque Based on Positional Convolutional Attention Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] Step a) comprises the following steps:

[0047] a-1) by formula The area of ​​vascular plaque in CT images after normalization processing was calculated by z-score normalization method where x original is the input vascular plaque CT sample, μ is the mean value of the batch data, σ is the variance of the batch data, and π is a given constant to prevent the denominator of the formula from being 0;

[0048] a-2) In the images obtained by computerized tomography scanning, the involuntary movement of the human body will cause artifacts in the detection results. Therefore, noise reduction processing is performed on the CT images of vascular plaques. Due to the difference of CT detection equipment, the size of the image will be different, so the normalized image It is m rows and n columns, through the formula will image Represented as a two-dimensional array, by the formula F(x,y)=median x,y∈around(x,y) [f(x,y)] uses median filtering to image Perform noise reduct...

Embodiment 2

[0051] In step c), the image D is processed by a two-dimensional convolutional layer and then processed by batch normalization and a Sigmoid activation function.

Embodiment 3

[0053] In step d) through the formula Calculate the feature map D 4 ,D 4 ∈ R Q×Q , where α is the scaling factor, T is the matrix transpose, is the feature map D 1 The i-th pixel in , is the feature map D 2 The jth pixel in , through the formula calculate and The degree of correlation and dependence of the locations, when the correlation between them is greater, the feature representations of the two locations are more similar.

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

A method for segmenting CT images of vascular plaques based on positional convolutional attention networks. The preprocessing operations of normalization, denoising, and enhancement are performed on the original collected images in sequence; secondly, the preprocessed images are input into Positional Convolutional Attention Networks for Segmentation. The positional convolutional attention network consists of two modules: the first module is the positional attention module, which uses the self-attention mechanism to capture the positional correlation and dependence between image pixels, so as to generate an aggregated image feature map, so that through the association The extremely high similarity shown by the high-degree pixels forms a preliminary segmentation area. Module 2 is the V‑Net structure, which uses downsampling, upsampling and skip connection structures to generate a restoration feature map. The restoration feature map incorporates more pixel features, making the information such as the edge of the segmentation map more refined; finally, the position convolution The aggregated image feature map output by the attention network is fused with the restored feature map to obtain the target segmentation image.

Description

technical field [0001] The invention relates to the technical field of medical imaging, in particular to a CT image segmentation method for vascular plaques based on a positional convolution attention network. Background technique [0002] The current mainstream vascular plaque screening technology is to use CT technology to form images, which are then manually analyzed. But the level of human analysis is often limited. In the past, medical image segmentation was mainly based on traditional image segmentation methods. The method is to simply use the texture, shape, grayscale and other characteristics of the image to divide the image into several disjoint regions, and there is no objective criterion for the segmentation performance. . With the development of artificial intelligence technology, automatic image analysis technology using deep learning can greatly improve the time and efficiency of analysis on the basis of replacing manpower. Therefore, the research on automati...

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 Patents(China)
IPC IPC(8): G06T7/00G06T7/12G06T5/00G06T5/50G06N3/04G06N3/08
CPCG06T7/0012G06T7/12G06T5/50G06N3/08G06T2207/10081G06T2207/20032G06T2207/20021G06T2207/20081G06T2207/20084G06T2207/30101G06T2207/20221G06N3/045G06T5/70
Inventor 王英龙徐鹏摇舒明雷周书旺
Owner SHANDONG ARTIFICIAL INTELLIGENCE INST