ACS auxiliary detection method for coronary artery CT radiography and computer readable storage medium

By combining generative adversarial networks and Unet networks, the independent problems of CT image denoising and super-resolution were solved, high-quality image reconstruction and plaque recognition were achieved, and the accuracy and efficiency of medical testing were improved.

CN120807283APending Publication Date: 2025-10-17HANGZHOU DIANZI UNIV +1
0 Cites 0 Cited by

Patent Information

Application Number
CN202510810055.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-06-17
Publication Date
2025-10-17

AI Technical Summary

Technical Problem

The existing CT image denoising and super-resolution tasks are usually performed independently, resulting in noise amplification or loss of high-frequency detail information, making it difficult to achieve high-quality image reconstruction at the same time.

Method used

A generative adversarial network combined with multi-task learning is used to achieve image denoising and super-resolution reconstruction through the generator and discriminator. The image quality is optimized by combining adversarial loss, perceptual loss and pixel-level loss functions, and the Unet network is used for plaque recognition and analysis.

Benefits of technology

It achieves the generation of high-quality images, improves the reliability of medical testing and the accuracy of plaque identification, and reduces the subjective errors in manual film reading.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to an ACS auxiliary detection method for coronary artery CT radiography and a computer readable storage medium, and the method comprises an image quality enhancement module which is used for carrying out the denoising and super-resolution reconstruction of an input CT image, and generating a high-quality image, and a plaque recognition module which is used for carrying out the recognition of the plaque. The technical scheme has the advantages that the high-resolution de-noised image is generated through the generator, effective features learned by the generator are increased through the discriminator, and high-efficiency image quality enhancement is achieved; by combining an adversarial loss function, a perceptual loss function, a total variation loss function and a pixel-level loss function, the image is close to a real reference image, and the reliability of medical detection is improved; through the first Unet network component, the second Unet network component and the double-attention module, the feature extraction quality is improved, and the subjective error of manual film reading is reduced.
Need to check novelty before this filing date? Find Prior Art