Image recognition method for predicting the risk of complications after cardiovascular interventional procedures

By analyzing the linear texture and branching structure of cardiovascular angiography images, artifact regions are screened out, solving the problem of low efficiency and accuracy of artifact recognition in existing technologies, and improving the accuracy of predicting the risk of complications after cardiovascular intervention.

CN121190543BActive Publication Date: 2026-03-10西安市人民医院(西安市第四医院)
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-09-23
Publication Date
2026-03-10

AI Technical Summary

Technical Problem

Existing technologies for identifying artifact regions in cardiovascular angiography images using target detection methods are inefficient and inaccurate, especially since band artifacts are difficult to identify due to their similarity to the tubular structures of the cardiovascular system.

Method used

By analyzing the linear texture distribution and branching structure of cardiovascular angiography images, combined with changes in vessel diameter and contrast agent fluctuations, edge detection and Hough line detection were used to screen for band artifacts. The neighborhood interpolation method was used to supplement the edges, and the tubular structure was extracted by combining the region growth algorithm and skeletonization technology. Anomaly evaluation indicators and vessel change parameters were calculated to screen out artifact structures.

Benefits of technology

It improves the accuracy and efficiency of artifact region identification, reduces the misidentification of band and tubular artifacts, and enhances the accuracy of predicting the risk of complications after cardiovascular interventional procedures.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121190543B_ABST
    Figure CN121190543B_ABST
Patent Text Reader

Abstract

This invention relates to the field of image artifact recognition technology, specifically to an image recognition method for predicting the risk of complications after cardiovascular interventional procedures. Firstly, based on physiological structural features, the method identifies band-shaped artifacts by recognizing the branching and tortuous characteristics of tubular structures corresponding to blood vessels, and the distinguishing features of regularly distributed, independently distributed band-shaped artifacts exhibiting high angular consistency. The method first filters out band-shaped artifacts based on the consistent distribution of straight-line textures. Then, by combining the complex branching structure of blood vessels, the temporal changes in blood vessel diameter, and the fluctuations in contrast agent with blood flow, the probability of artifact structures is calculated, thereby filtering out tubular artifact structures with lower conformity to the aforementioned blood vessel features. This results in more accurate identification of band-shaped and tubular artifact structures representing the artifact region, and compared to target detection methods, the machine learning-based image recognition processing method is more efficient.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Method and device for artifact detection

    CN102018524A

  • Systems for detecting and tracking of objects and co-registration

    US20160196666A1