Pet abdomen ultrasonic detection method and system based on image analysis

By processing pet abdominal ultrasound images in the frequency domain, Fourier transform and Gaussian filtering techniques are used to accurately reduce hair root scattering noise, solving the problem of distinguishing noise from the edges of small lesions and improving the diagnostic accuracy of pet abdominal ultrasound examination.

CN122368068APending Publication Date: 2026-07-10NANJING AGRICULTURAL UNIVERSITY
0 Cites 0 Cited by

Patent Information

Application Number
CN202610831007.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-10
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing technologies cannot effectively distinguish between structured stripe noise caused by residual hair root scattering and weak edges of small lesions in organs, resulting in damage to anatomical features while suppressing noise and reducing the accuracy of lesion identification.

Method used

By mapping abdominal ultrasound images to the complex frequency domain, using two-dimensional fast Fourier transform and the autocorrelation coefficient of the high-frequency phase value sequence, the feature orientation angle is locked, a two-dimensional Gaussian filter mask is constructed for element-wise multiplication, and the edge suppression feature map is reconstructed through two-dimensional inverse Fourier transform. The data is then input into a machine learning model for assisted diagnosis.

Benefits of technology

It achieves precise reduction of structured physical noise, preserves weak edge features of organs, reduces false alarm rate and false negative rate, and improves the diagnostic accuracy of abdominal ultrasound examination in pets.

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

Abstract

The invention discloses a pet abdomen ultrasonic detection method and system based on image analysis, and relates to the technical field of pet abdomen ultrasonic image processing.The method comprises the steps that an ultrasonic image is mapped to a complex frequency domain space through two-dimensional fast Fourier transform, and a space phase spectrum is extracted; taking a zero frequency center point as an origin of a radiation point, filtering out a low-frequency component of a substantial organ contour, traversing non-axial and radial sampling rays in a preset high-frequency band, extracting a phase value sequence, calculating an autocorrelation coefficient, and retrieving a characteristic direction angle corresponding to a maximum value. And constructing a two-dimensional Gaussian filtering mask, and multiplying the two-dimensional Gaussian filtering mask by a two-dimensional complex frequency domain matrix element by element so as to specifically reduce the physical grid stripe noise in the direction. And finally, restoring to an airspace through two-dimensional inverse Fourier transform and reconstructing to obtain an edge suppression feature map, and inputting a machine learning model to output a focus auxiliary probability report. According to the method, structured physical grid noise caused by pet residual hair roots is specifically filtered out, and the focus recognition accuracy is remarkably improved.
Need to check novelty before this filing date? Find Prior Art