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

A Nafld Ultrasound Video Diagnosis System Based on Siamese Attention Network

A diagnostic system and attention technology, applied in the field of short video processing, can solve the problems of irrelevant information interference, low-quality ultrasound features, etc., to achieve the effect of solving irrelevant information interference, alleviating negative effects, and improving performance

Active Publication Date: 2022-05-10
XIAMEN UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] For the task of NAFLD diagnosis in ultrasound video, the main problem is the interference of irrelevant information and the problem of weak features caused by the low quality of ultrasound itself

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 Nafld Ultrasound Video Diagnosis System Based on Siamese Attention Network
  • A Nafld Ultrasound Video Diagnosis System Based on Siamese Attention Network
  • A Nafld Ultrasound Video Diagnosis System Based on Siamese Attention Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0043] The invention discloses a NAFLD ultrasonic video diagnosis system based on a twin attention network, which consists of two twin attention sub-networks with the same structure and weight sharing, and the twin attention network consists of two twin attention sub-networks with the same structure and weight sharing Attention sub-network and loss function composition, such as figure 1 As shown, the Siamese attention sub-network consists of two-stream feature extraction module a, linear classification module b and contextual attention c module, and the loss function consists of binary cross-entropy loss, contrastive similarity loss and contrastive difference loss.

[0044] Wherein, the dual-stream feature extraction module a includes a sharing module, a classification module and an attention module; the dual-stream feature extraction module a is used to extract different features of classification and attention.

[0045] The sharing module is used to extract the low-level fea...

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 NAFLD ultrasonic video diagnosis system based on a twin attention network. The system consists of two twin attention sub-networks with the same structure and shared weights and a loss function, wherein the twin attention sub-network is extracted by dual-stream features. module, linear classification module and contextual attention module, and the loss function consists of binary cross entropy loss (BCE), contrastive similarity loss (CSL) and contrastive dissimilarity loss (CDL). The invention adds a dual-stream feature extraction module on the basis of the twin attention network, and introduces a loss function, so that the NAFLD ultrasonic video diagnosis system achieves an accuracy rate of 90.56%, a specificity of 88.26%, and a sensitivity of 93.58%, which is Ultrasound video diagnosis of NAFLD provides an efficient and feasible method.

Description

technical field [0001] The invention relates to the technical field of short video processing, in particular to a NAFLD ultrasonic video diagnosis system based on a twin attention network. Background technique [0002] Early screening of nonalcoholic fatty liver disease (NAFLD) helps patients prevent irreversible advanced liver disease, but manual diagnosis of ultrasound videos of NAFLD requires doctors to browse lengthy videos, which is tedious and time-consuming in clinical practice. Therefore, the method of deep learning can be used to realize the automatic diagnosis of NAFLD in ultrasound video to improve the efficiency of diagnosis. [0003] For the task of NAFLD diagnosis in ultrasound video, the main problems faced are the interference of irrelevant information and the problem of weak features caused by the low quality of ultrasound itself. Contents of the invention [0004] In order to solve the above problems, the present invention provides a NAFLD ultrasonic vid...

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): G06T7/00G06V10/764G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06T7/0012G06N3/049G06N3/08G06T2207/10016G06T2207/10132G06T2207/30056G06T2207/20081G06T2207/20084G06N3/048G06N3/044G06N3/045G06F18/2451
Inventor 王连生
Owner XIAMEN UNIV